The Theory and Practice of Rx Profiling in Mortality And Morbidity Risk Underwriting
I. Prologue The Purpose Of This White Paper
II. Perceived And Proven Value Of Rx Profiling In Life And Health Underwriting
- Top Ten Underwriting Payoffs From Rx Profiling Plus The Rx Profiling Experiences Of Two Companies
- How One Insurer Is Realizing Direct Payoffs From Rx Profiling
- The Rx Profiling Experience Of A Major Health Insurer
III. Essential Details About Rx Profile Services
IV. The Implications Of Nondisclosure
V. Off-Label/Unapproved Rx Prescribing
VI. Patient Adherence To Prescribed Drugs
VII. The X Factor
VIII. Rx Profiling And Teleunderwriting
IX The Protective Value Of Rx Profiling
X. Recommendations As To How To Interface Rx Profiling With Your Underwriting Process
XI. Report Of A 52 Company Rx Profiling Survey
- Part One: Perceptions
- Part Two: Experiences With Rx Profiling
- Section A
- Section C
- Section D
- Survey Comments
XII. Closing Comments
As mortality and morbidity underwriting undergoes a much-anticipated metamorphosis to accommodate the demands of the 21st century, the need for fast, reliable and cost-effective screening resources has come into clear focus. The purpose of this paper is to examine the extent to which Rx profiling is one of these resources.
Knowing which pharmaceuticals an insurable applicant is taking has always been a key goal of medical history-taking. It is axiomatic that if we know what Rx an individual is taking, we have keen insight into his medical history. It is an a priori assumption, therefore, that Rx profiling services will serve to enrich our ability to relate pharmaceuticals to medical histories.
We identify ten ways in which Rx profiling impacts the insurers bottom line. Some are now seen as universal mandates (e.g., speeding up the underwriting process, reducing medical record dependency, and improving mortality and morbidity outcomes, reducing nondisclosure/antiselection). Others are more subtle. These include confirming the completeness and accuracy of Rx information admitted during the application process, allowing the underwriter to affirm stipulated diagnoses and also pinpoint comorbid conditions that may not be disclosed, and assuring Rx applicant compliance with prescribed drugs.
One proactive life insurer is in the midst of a protect designed to assess the value of Rx profiling as a way of both reducing their dependence on age/amount medical records and increasing their success in detecting material nondisclosure. At this writing, both goals are being realized and the expectation is that they will be able to further extend the use of Rx profiles for larger face amounts and at older ages.
Key direct contributions of Rx profiling include the rising percentage of applicants for whom a record is found on the database (65%, a major improvement over just a few years ago), the sometimes-underappreciated impact (as we show by example) of medication dosage and times-taken, and the ability to see the whole range of prescriptions filled for greater insight into the nature of underlying impairments, and so on.
A long-held misperception is that nondisclosure is driven largely by willful antiselection. It is now appreciated that 90% of nondisclosures are, in fact, incidental, and that their prevalence is due mainly to inadequacies in how insurers gather core medical history information. Rx use stands out prominently in this regard, as the lay applicant struggles to clearly name, let alone amplify, details of current and recent medication use. Rx profiling works to counter incidental antiselection.
It is said that one in every four prescriptions is written off-label; that is, for medically-sound reasons that are not as yet approved by the Food and Drug Administration (FDA). Through enhancing the basic Rx profile with a new and unique adjuvant resource to be know as Hanks Underwriting Guide to Rx; due in mid-2006 the underwriters ability to correlate unapproved medication use with Rx profile findings will be assured.
Insurers are increasingly willing to grant preferred and even super-preferred coverage to individuals taking antihypertensive and lipid-lowering drugs. Non-adherence to Rx is a major issue in clinical medicine and it would have a formidably deterimental impact on this practice. Therefore, one of the immediate-impact advantages of the Rx profile is that it confirms if, indeed, the individual is refilling his prescription.
There is growing recognition that wide range of health habit/lifestyle choices directly impact insurability. For example, if one is compliant withRx use, this is likely an indicator that the individual is also disposed to abide by M.D. recommendations for periodic screening (Pap smear, mammogram, stool occult blood and so on). This is another, albeit indirect, way in which confirmation of Rx adherence impacts the determination of risk.
Teleunderwriting is fast becoming the dominant mode of medical history-taking, replacing less-through histories non-medicals completed by producers, as well as those taken in conjunction with paramedical and medical examinations. Teleunderwriting is defined as the use of telephone interviews, with drilldown questioning of all YES answers about medical history, as a surrogate for these traditional modes of history-taking.
The ultimate success of teleunderwriting depends on the underwriter having as much information about the proposed insured as possible. To this end, Rx profiling makes a major positive contribution because it amplifies the relative importance of medical history information given by the lay informant. It is clear now that Rx profiling is one of a handful of feasible screening tools that will make or break the impact of teleunderwriting on risk management.
There is a definite protocol that optimizes the capacity of the insurer to evaluate Rx profiling. The same is true when it comes to directly interfacing this resource with the prevailing risk selection regimen. These issues are discussed in this paper.
Finally, we report on a 52-company survey on Rx profiling. This unpublished survey completed in October/November, 2005 - inquiries at length as to perceptions about the pros and cons of the Rx profile, while also collecting objective information regarding the experiences of those insurers who have already reviewed and, in many cases, embraced Rx profiling in day-to-day underwriting.
It is notable that these survey results show us how one-time concerns about potential obstacles to Rx profiling have largely disappeared and moreover that today most insurers are objectively assessing whether or not Rx profiling works in their underwriting environment. A review of this survey will be instructive to anyone seeking to gauge the viewpoints and experiences of a broad range of life and health insurers as regards the theory and practice of Rx profiling.
I. Prologue: The Purpose of This White Paper
This underwriter agreed to undertake this white paper project first and foremost because of an abiding interest in carefully examining the value of Rx profiling services in life and health underwriting.
When the early versions of this service appeared in the mid-to-late 1990s, I confess to being more than a little skeptical about both their suitability and their potential value as risk assessment resources. Concerns about how this practice would be received by producers, regulators and the public at large loomed ominously. At the same time, there was the issue about whether or not deriving Rx use information in this manner would produce sufficient protective information to justify the costs. These issues, and many others, will be addressed and discussed in this white paper.
The emergence of teleunderwriting as the increasingly-dominant mode of risk appraisal information-gathering has had a strong and positive impact on the value of Rx profiling. Why this is true and what its implications are for underwriting in the years ahead will be examined as well.
My initial conceptualization of the ways in which Rx profiling could impact risk selection as well as other aspects of life and health operations, most notably claims, was driven by the prospect of detecting nondisclosure. It was only with time, and feedback from members of my five life and health underwriting study groups and others that I came to appreciate the full range of potential pay-offs from the information gathered via Rx profiling. These pay-offs will be outlined and discussed at length in this paper.
One of the best ways to discover industry perceptions about almost anything related to underwriting is to do a survey of chief underwriters. Accordingly, I undertook such a survey on Rx profiling. In it, I sought both subjective impressions as well as objective responses to questions regarding experimentation with and current use of Rx profiling. Full details of this survey are reported in the final section of this white paper, along with summarized commentary from both survey respondents and additional observations by this underwriter.
It is my hope that this white paper will assist life, health, disability, long term care and critical illness insurers in the US market in further evaluating the pros and cons of Rx profiling for potential use in their companies. Perhaps this paper will also generate further discussions among underwriting professionals as to the ways that deployment of Rx profiling as a screening requirement could contribute to faster and more effective risk management.
I encourage readers to contact me to discuss the topics and issues raised in this paper. Because of my extensive travel, I am best reached via email at firstname.lastname@example.org.
My sincere thanks to Ingenix MedPoint for engaging my services to research and write this white paper. This prodigious undertaking could not have been accomplished without their assistance and support.
I would like to make it clear that what is written herein represents solely the research, analysis and opinions of this author. My sponsor and I had a clear understanding that what I write is to be wholly objective, rather than simply a commercial message.
One of the addenda to this paper is devoted specifically to marketing information regarding the specific value of Ingenix MedPoint. The other highlights the features of my own forthcoming Rx profiling support resource, to be known as Hanks Underwriting Guide to Rx.
These subjective promotional elements are strictly confined to their respective addenda and are not related to the research, findings and conclusions reported in the body of this white paper.
Hank George, FALU, CLU, FLMI
Hank George, Inc.
December 1, 2005
II. Perceived and Proven Value of Rx Profiling in Life and Health Underwriting
"Tell me what hes taking and Ill tell you what hes got! "
Hank George, FALU, CLU, FLMI
Northwestern Mutual agency visit
How well I remember the first time I uttered this statement. It was my first time visiting and speaking at a general agency on behalf of my first industry employer, Northwestern Mutual.
I brought up Rx in the context of urging these producers to do everything they could to make their clients medical histories clear to underwriters, using what I called the five essential questions that must be asked and answered about any medical impairment.
I was trying to make a point about carefully documenting how their clients were treated, so as to maximize the amount of risk-relevant information at the underwriters disposal and thus minimize the delay in getting their business approved and issued.
I used a 35 mm slide to jog my memory as I spoke. The slide related to the question of Rx information read: Rx = Dx (a cryptic way of saying the same thing as the above quote).
I dont know if I had any genuine impact that day, but I do know that this phrase as stayed with me for almost 30 years. Indeed, whenever I address a producer audience anywhere in the world on how to get their business moved through underwriting, one of the points I hammer home is the incredible value to underwriters of knowing precisely how their client is or was treated in order to better assess the mortality risk while avoiding potentially unnecessary requirements.
This, of course, is music to their ears!
Top 10 Underwriting Payoffs from Rx Profiling
Plus The Rx Profiling Experiences of Two Companies
As I reviewed hundreds of pages of notes from study group meetings, survey responses, published articles and other sources where the issue of Rx and Rx profiling has come up in a risk appraisal context, I had to discipline myself to hone down the information I would present here to make this white paper palatable to the reader.
With regard to the return from Rx profiling, I did this by focusing on what looked to be, dollar for dollar, the TOP TEN returns for the underwriting process if Rx profile information is included in the companys screening requirement protocol.
I will list these in order of perceived impact on the risk management process, citing references and making appropriate comments as I do so:
#1 Speeding Up the Underwriting Process
"[Rx profiling] has helped us save time and money, and the risk is that much better."
Kathy Kirk, Trustmark Insurance
TechEdge: Insurers Put New Products to the Test
Tech Decisions for Insurance
The National Underwriter Company
Speeding up the underwriting process is most commonly measured in terms of reducing application cycle time. Another term for cycle time is application-to-issue time. Either way, we are talking about how long it takes to issue the average policy, based on the interval that commences when the new business department receives the application for insurance and ends when an underwriting decision is taken and, hopefully, a policy is issued.
Historically, underwriting turnaround time has been ponderous. The primary culprit in this regard has mainly been due to the time the application spends in the underwriters pending case file. That interval is largely determined by how long it takes to gather in requirements. The slowest requirement and the one whose tardiness has been most difficult to remedy is the APS (attending physicians statement).
How has slow turnaround time been perceived by the distribution side?
A 2003 Independent Producer Survey undertaken by Celent Communications, a Boston-based consulting firm, was reviewed in the March, 2004 issue of Bests Review by Ron Panko, one of the senior members of that publications editorial team.
Ron listed the factors, reported in the study, that producers ranked according to how those factors impacted their choice of the carriers through whom to place their new business. The second most significant factor, based on the percentage of respondents who identified it as very important was carriers ability to deliver policies quickly. It is interesting that 55% of those who completed the survey ranked application-to-issue turnaround time as very important, whereas the element we traditionally believe to be at the top of their wish list commission rates paid on their new business actually finished third, with a 52% very important rating!
A 2004 LIMRA report titled Insights Into Strengthening Producer Relationships by Denise C. Marvel, CLU, ChFC, FLMI, ACS, similarly underscored the perceived importance by the folks who bring in the revenue of fast and competitive underwriting. This was the third most mentioned topic when Ms. Marvel surveyed a cohort of producers regarding ways to carriers to improve producer productivity. Clearly, accomplishing this mighty feat (greater producer sales volume) would be a top priority for insurers as well.
One critical impact of application-to-approval turnaround time is on the insurers not-taken rate. This is the percentage of potential business that is ultimately not placed in force due to either client loss of interest or the success of a competitor. What drives not-takens is sluggish cycle time.
According to Ron Verzone, president of United Underwriters, Inc. and a frequent and respected contributor to many publications, cycle time is a key consideration in this regard. Writing in the October, 2003 issue of Bests Review, Ron states that by reducing the approval process by just a couple of weeks, we could cut the default (that is, not placed) figures in half.
This message has been reinforced many times by underwriters as well, which is understandable considering that they often bear the brunt of producers angst associated with long-pending business.
One workflow process consultant has summed this up succinctly:
"If the underwriting cycle is reduced by five days, it is reasonable to project reducing the not-taken ratio by 8% - 10%."
On The Risk
How does turnaround time relate to the Rx profile?
Being accessible to the underwriter almost instantly, the Rx profile is one of a handful of requirements that directly facilitate rapid cycle time. This is in stark contrast to certain traditional medical underwriting resources such as treadmill stress tests and chest x-rays, which are notable for their adverse impact in this regard.
Moreover, the Rx profile also contributes dramatically to reducing the insurers volume of APS ordering, which is the point we will now review as the second of the 10 pay-offs from Rx profiling.
#2 Reducing Insurer Overdependence on Medical Records
"The biggest news in 2001 is the availability of databases that track the use of prescription medications. Potentially, this can reveal all serious medical ailments known to the applicant. Not only does this have some potential as an APS replacement, but it also can be used to corroborate and supplement information provided on the application."
Maria Thomson, FSA
Thomson Management Solutions, Inc.
The National Underwriter
July 23, 2001, pages 12 and 14
In recent years, this underwriter has spoken often about the APS over-dependency syndrome as one of the critical issues facing insurers.
Although the impact on cycle time is a dominant reason for characterizing this phenomenon as a syndrome, the dimensions of the problem extend beyond this consideration. Recent regulatory and legislative actions are widely expected to exert an ultimately negative effect on medical record acquisition, both in terms of the quantity and quality of medical history information available to the underwriter. Accordingly, finding a way to reduce our need for these medical records ranks high among the priorities of mortality and morbidity risk chief underwriters.
Kathy Kirk of Trustmark Life sums up the impact of Rx profiling in this regard in the earlier-cited publication, saying if you know what kind of medication [customers] are taking and how often they are taking it, we dont need the medical records as often.
This works in two ways.
First, it is reasonable to postulate that Rx profiles either showing that no Rx given or that the medicine prescribed is consistent with an insignificant condition - often validates the fact that the routine medical check-up or some other vague or nonspecific MD visit revealed nothing of mortality or morbidity significance.
"Not many people who walk out of a physicians office with a significant symptom or diagnosis do not have, in hand, a prescription it has almost become an expectation."
Hank George, FALU, FLMI, CLU
Rx = Dx: Pharmacology and Underwriting
Proceedings of the 2001 Annual Meeting of the Institute of Home Office Underwriters, page 177
This reality relates directly to the issue at hand.
If a person sees his physician with almost any of symptom or other medical complaint, and that symptom or complaint speaks to the presence of an underlying medical condition, the likelihood that the individual will exit the physicians office with one or more medication prescriptions to fill and take is very high. Therefore, if Rx is not given, that symptom or complaint is apt to have been insignificant and thus of little or no concern to the underwriter.
This is not true in every situation, of course. For example, an enlarging lymph node may be the first sign of cancer and the usual approach in this case is either biopsy or a period of observation, depending on the context of the situation. Rx is usually not prescribed in this situation and this fact does not lessen the importance of explained lymph node enlargement. But, this and equivalent scenarios are the exception and should be readily recognized as such by an experienced underwriter. They do not diminish the validity of the statement that, on the vast majority of occasions, the fact that Rx is not prescribed is a very strong indicator of little or no risk associated with the symptom or complaint under consideration for insurability
Second, when the risk implications of an impairment is vague (e.g., chest pain, headaches, fainting), the Rx prescribed may point convincingly to a favorable diagnosis and thereby allow for approval without further underwriting.
A fitting example here is joint pain (arthralgia). If after evaluating this complaint, treatment is limited to analgesics, the arthralgia is likely due to some self-limited cause, such as trauma. On the other hand, if the physician prescribes methotrexate, an experienced underwriter will know immediately that rheumatoid arthritis, or some other serious joint disorder, is present.
In either the first or second scenario discussed above, an APS that would otherwise have been ordered is not pursued because of what is learned, in large measure, via the Rx profile of the proposed insured.
As I hope you see, intrinsic to the ability of an underwriter to make accurate inferences as to underlying impairments is his awareness of all medications taken. Unfortunately, however, few underwriters come into the profession with a background in pharmacology Therefore, they need to not only increase their knowledge base on this subject but also to have rapidly-accessible resources that can assist them in sorting out these cases. We will revisit this subject in Section VIII on teleunderwriting.
For now, suffice to say that any rapidly-acquired and affordable requirement can favorably impacts a reduction in the volume of APS orders is going to attract a lot of positive attention among chief underwriters (and their bosses)!
#3 Improve Mortality and Morbidity Outcomes
As underwriting continues its metamorphosis from the ways of the 20th century to accommodate the demands and priorities of the new millennium, all innovative changes must satisfy the mandate to at least sustain, if not improve, mortality and morbidity outcomes. This is the bottom line where underwriting is concerned.
This is becoming a senior-management imperative because of prospects for greater mortality and morbidity gains in other words, profits that depend in large measure upon accurately underwriting risks.
It comes down to this: protective value sacrificed to allow for modernization must be compensated for by additional protective value from resources. It is here that the contribution of Rx profiling is coming into clearer focus.
To have the impact that is needed, Rx profiling needs to become an integral part of the underwriting screening model. In this way, its contribution to aggregating sufficient protective information will be maximized. This is because no matter how much value any one new asset brings to the table to offset what must be dispensed with for the sake of progress, it is the collective return that ultimately determined where or not mortality and morbidity outcomes will be optimized.
These are the potential resources that insurers can realistically consider for building their new paradigms. Some of these are better suited for mortality, others perhaps for morbidity. But the net effect has to be that the payoff hits the mark:
- Rx profiling which gives every indication of being a co-anchor of this process.
- MIB the other co-anchor.
- Oral fluid screening which is now established at ages 18-39 under an ever-broadening threshold.
- Motor vehicle record (MVR) long regarded as a screening resource only in males, ages 18-30, we now see that universal MVR acquisition pays off handsomely in both genders and at all ages.
- Skin cholesterol testing this novel alternative coronary disease risk predictor is being field-tested at this writing. Its potential is obvious, given that it targets cholesterol deposition and can be done in one minute by an insurance agent
While this package of screening tools works well in a conventional underwriting shop driven by non-medical, paramedical and MD exam history-taking, it is even more effective where teleunderwriting has been adopted and the volume of APS ordering minimized.
On balance, one is struck as regards achieving mortality and morbidity goals - by what can be accomplished simply by integrating Rx profiling into ones risk management model.
#4 Confirming the Completeness and Accuracy of Rx Information Derived from Application Part IIs, Paramedical and other Examinations, Teleinterviews and MD reports
The Rx profile is not the only source of medication use information. However, what separates it from the others is its higher degree of completeness and accuracy of content.
It is hardly a secret that Rx information gathered by producers is usually replete with misspelled drug names and omissions of some (in many cases, all) medications. Unfortunately, the magnitude of improvement here by using paramedicals and MD examinations (to the extent the latter are still being used) is not nearly as great as one might expect.
Teleinterviewing done properly is, by consensus, the most successful mode of information-taking when it comes to the amount and quality of Rx information. However, even with accurate drill-downs of medical histories, experience teaches that we do not get consistently complete and accurate medication information.
Given the shortfalls of other modes of obtaining reliable information on Rx use, it is easy to see why Rx profiling confers such potentially great value for insurers
#5 Nondisclosure and Antiselection
The great nemesis of underwriters is nondisclosure of significant information by the insurance seeker. This may be inadvertent or done with intention, as we shall discuss further on in this white paper.
Most chief underwriters, in any product line of mortality and/or morbidity insurance, would most certainly concur that any affordable and appropriate underwriting resource that would help them to pinpoint and then intervene on nondisclosure would be worthy of their careful consideration as a risk appraisal asset.
The value of the Rx profile in this context is obvious. In fact, it is the prospect of uncovering nondisclosure that first attracted insurers to Rx profiling.
Paradoxically perhaps, the perceived failure of Rx profiles to identify a sufficient number of such nondisclosures is the main reason why some insurers who undertake a trial with one of the Rx resource providers decline to purchase access to Rx profile information on an ongoing basis.
The bottom line here is that Rx information does alert us to some instances of nondisclosure, and in many cases this nondisclosure is material and would have changed the underwriting outcome but, again, this is just one of the payoffs from Rx profiling and it should never be perceived as the sole determinant of its value to a given carrier.
#6 Expanding the Underwriters Capacity to Pinpoint Clinical Diagnoses
Experience shows us that many insurance applicants do not have as much insight into medical diagnoses as we would wish. The major antidote for such uncertainty is (unfortunately, in consideration of its aforementioned drawbacks) the physicians statement.
In teleunderwriting, while we get a great deal of salient risk information from the applicant, experience teaches that asking the proposed insured what diagnosis he was given does not often yield sufficient credible information with which to frame an underwriting decision.
Rx information allows us to greatly reduce our determination of the most probable diagnosis, because medications are given for a finite number of reasons.
For example, if a person has chest discomfort and is prescribed a nitrate vasodilator, the diagnosis is almost invariably ischemic heart disease.
Conversely, if an anti-reflux Rx is given, coronary disease is improbable and our focus shifts to esophageal reflux.
The main obstacle to pinpointing a diagnosis using Rx information is the growing incidence of off-label prescribing by physicians. Off-label refers to prescribing medications that are not yet approved by the Food and Drug Administration.
At the present time, I am not aware of any resources that consistently and accurately identify potential unapproved uses for most medications. This problem will be dealt with definitively with the advent of my Rx resource. When this happens, integrating Rx profiling reports, all other known medical history and the results of lab tests will be a much smoother and more accurate process.
I will provide some examples of this dilemma further along when we discuss the issue of off-label/unapproved Rx prescribing in more detail.
#7 Facilitate the Jet Issue Process and Simplified Issue Underwriting
These two very different subjects are considered together because the end point in underwriting is to take action on as many cases as possible as rapidly as possible and with as little protective information as possible.
Jet underwriting may be used for 30% or more of new business in a typical life or health carrier. If the case is clean (defined as only trivial conditions or no medical history whatsoever), it is approved automatically, without being seen by an underwriter.
Matching an Rx profile to a clean case can be very revealing. It may be the only evidence, other than occasional unexpected MIB codes, of an undisclosed impairment.
Using Rx profiling in this regard allows insurers to cut out a small but very significant number of cases, which would otherwise have been inappropriately issued as applied for. This assures even better mortality and morbidity outcomes from the remaining jet-issued cases in that block of business.
"Fast turnaround time is one of the cardinal selling points of [simplified issue]. Swift Underwriting decisions are expected and demanded."
Simplified Issue Requires a Different Mindset
Cary Lakenbach, FSA, MAAA, CLU
Actuarial Strategies, Inc., Bloomfield, CT
The National Underwriter
December 6, 2004, page 16
Simplified issue underwriting has now been configured for almost every product in life, healthy and disability insurance. The core concept is that only a few risk-related questions are asked (five or even fewer for the most part, focusing mainly on major impairments). A YES answer to one of these so-called knock out questions will either result in immediate declination or, in some cases, further underwriting. A NO to all of them assures immediate approval.
The knock out questions are usually broadly worded and for this reason inherently lead to non-disclosure when the proposed insured perceives an impairment differently from how it is described.
A good example is acute coronary syndrome (ACS), a relatively new cardiology term encompassing the distinct entities: transmural (full wall thickness) MI, subendocardial (affecting only a portion of the heart muscle wall) MI and unstable angina (which is not strictly-speaking a heart attack because it does not result in proven heart muscle death).
If the patient is told that he had an ACS event or that he had something other than a heart attack, and if heart attack is the all-encompassing single term used on a simplified issue application, then there is a good chance that the proposed insured will answer NO. But if Rx information consistent with ACS is disclosed on the Rx profile, this will alert the underwriting to pursue more medical history details before accepting the risk.
Even a few Rx profile positives could have a profound implication on profitability of simplified business, not only because of situations like the one just described but also because asking about medication use is rarely one of the knock out questions.
#8 Matching Impairments to MIB codes
MIB codes are one of our most cost-effective protective value assets. They provide a clue to a possibly significant undisclosed impairment.
Whether to pursue more information in these situations depends on what we know from the application and our initial set of screening information. If the coded impairment is not acknowledged but a telltale medication correlating with the code is revealed on the profile, the underwriter is better positioned to decide how far to pursue the situation.
#9 Identifying Comorbid Conditions
Comorbidities are impairments that co-exist with the primary disorder under consideration. They often have a huge effect on insurability. And, they are highly prevalent in most domains of medicine, especially psychiatric illness, chronic forms of circulatory disease and substance abuse disorders.
A recent study of 1137 depressed patients by Zimmerman et al. [American Journal of Psychiatry. 161,7(July, 2004):1285-89] found that 45.6% had at least one comorbid condition. Undoubtedly many of those comorbidities would have an impact on mortality and even more so on morbidity. The best clue to their presence will often be the use of a drug not routinely given in the context of depression.
An applicant may, for example, acknowledge a history of coronary disease or a past MI, but not disclose again, often incidentally the fact that during his medical evaluation for CAD he was also found to have an abnormally-low ankle-brachial index (which essentially confirms PAD - peripheral arterial disease). This is highly significant to insurability because the risk attributable to a combination of CAD and PAD is much more than merely additive.
The use of a drug for peripheral arterial disease, however, will alert the knowledgeable underwriter to the fact that there is more to the history than just chronic CAD or a past MI. In this way, Rx profiles routinely RED FLAG comorbidities that might otherwise go undiscovered.
#10 Adherence to Rx
Rx adherence or, if you wish, compliance is probably the most underappreciated area in which Rx profiling contributes to risk assessment. Therefore, we will consider this issue in considerable depth in Section IV.
How one insurer is realizing direct payoffs from Rx profiling:
One company that is currently exploring the value of Rx profile screening is Hartford Life, at their Minneapolis individual life office. According to National Underwriter OnLine breaking news on November 9, Hartford Life has announced that:
It will use prescription database checks rather than attending physician statements to underwrite applicants age 40 an younger who are applying for $2 million in deaths benefits and most applicants ages 41 to 60 who are applying for less than $1 million in death benefits.
National Underwriter reporter Trevor Thomas wrote a more detailed report on this project in the November 14, 2005 issue of that weekly magazine, citing Hartford Lifes Vice President and Chief Underwriter Chris Graham as saying that this initiative is part of a company plan to make it easier to buy life insurance.
To get more perspective on this groundbreaking approach to Rx profile screening, I interviewed an old friend, Dave Redpath, FALU, Underwriting Director at Hartford Life. Dave confirmed that this experiment has been going on for two months and that based on highly favorable findings thus far, it is continuing.
Dave said that the primary focus here was to reduce their dependence on age/amount APS reports (ordered, based on applicant age and amount of coverage when the proposed insured reports having seen a physician on a routine basis only, with no acknowledged significant medical history. Dave reported that they initiated this project because they found that 95% of such reports did not change the final underwriting decision.
When they relied on Rx profiles in lieu of routinely pursuing many of these APSs, they discovered they could eliminate over 50% of them, resulting in both significant savings and much faster issue of these policies.
It is helpful to note here that age/amount medical records are commonly acquired industry-wide. In fact, in survey undertaken by this underwriter several years ago, many life insurers reported obtaining more age/amount-driven APSs than they did medical records ordered for cause (defined as those procured because of a specific medical history reported by the proposed insured). With the ever-widening embrace of teleunderwriting, we are now able to amplify the circumstances surrounding what are said to be routine MD visits. By adding Rx profile records, teleinterview findings can be readily affirmed and favorable action taken.
It is also noteworthy that Hartford Life uncovered significant specific Rx details on 2 of the first 44 cases that had pharmaceutical records. In each of these, the facts disclosed on the Rx profile report changed the underwriting outcome by as much as much as 4 tables of extra mortality.
Dave advised that he will continue this project until at least 600 files have been reviewed and analyzed. Based on these initial findings, he hopes to present to the companys actuaries and their reinsurers a solid argument for expanding Rx profile screening, and to continue to reduce age and amount APS-ordering at even older ages and higher face amounts. We see here an example of how Rx profiling contributes directly to realizing two primary goals intrinsic to underwriting: ordering only those requirements that are truly needed and uncovering scenarios of significant nondisclosure.
The Rx Profiling Experience of a Major Health Insurer
The use of Rx profiling is escalating among individual and small group health insurers but little has been published to date on this subject related to the experience of proactive companies in this market.
I had the opportunity to interview the chief individual risk underwriting officer of a major Midwestern health carrier that has been using Rx profiling for just over 12 months, doing business in 20 states. I have permission from that chief underwriter to share these details regarding their experience to date.
Over the 10 month period ending October 31, 2005, this insurer conducted a volume of Rx profiles into the mid-five figure range.
Of those applicants who had Rx records, just over 93% of the Rx reports did not significantly alter the underwriting outcome of the cases. However, the remaining almost 7% did and this is a breakdown of their net impact:
- 3.3% resulted in avoiding the expense and delay of ordering an APS where it was no longer needed because of Rx detail
- 1.3% resulted in actually ordering an APS where one was needed and would otherwise not have been pursued
- Almost 2% identified an undisclosed drug that was significant to insurability
- 0.5% made a key contribution to deciding whether or not a paramedical examination was needed
All told, Rx profile findings were risk-significant led directly to 54% of applicants having an exclusion rider added, 31% of them being rated up and 15% being declined for coverage.
An internal cost-benefit analysis at this insurer proved that Rx profiling was definitely worth the money and this insurer continues to secure Rx profiles on all individually-underwritten health applications.
The chief underwriter added that the found considerable collateral value in having the names and contact information on physicians, especially in scenarios where the Rx data showed potential misrepresentation of the applicants true medical history.
This section reveals both the diversity and importance of the many ways in which Rx profiling contributes protective value in the mortality and morbidity underwriting process, as well as late-breaking details on an experiment at Hartford Life that is proving the impact of Rx profiles
My independent investigation of these ten payoffs from Rx profiling, as well as others that are implicit from what follows in this white paper, have convinced me that Rx profiles should now be considered a primary screening requirement for all individually-underwritten products encompassed under the broad term life and health insurance.
III. Essential Details About Rx Profile Services
The purpose of this section of the white paper is to use a question and answer format to cover the basic facts about Rx profiles that will assist the reader in fully understanding the issues discussed herein.
Some of these details will vary by provider. Since Ingenix is not only the sponsor of this white paper but also the largest provider of Rx profiling services in the industry, I have chosen to use as needed details related specifically to their product, known as MedPoint.
What are the sources of Rx data that are used to provide Rx profiles?
While the largest providers are pharmacy benefit managers (PBMs), medical clearinghouses, as well as data from pharmacy chains and health plans, are also accessed in this process.
Considering this diversity of sources, what is the lag time from the time of a request for an Rx profile until those data are provided to the insurer?
With Ingenix, access is described as near real time, meaning that the results are provided within minutes. This is extremely significant because our capacity to optimize rapid turnaround hinges on how soon the underwriter has the components of the risk screening process in hand to evaluate and to take action.
What bits of information are needed about a proposed insured to request an Rx profile report?
As one would expect:
- First and last name
- Date of birth
- Social Security number
- Zip code
Fortunately, all of these components are routinely gathered during the application process.
Is a signed HIPAA authorization needed?
Yes and fortunately so, as we want to strictly comply with all privacy and confidentiality regulations in place to protect personal health information.
On what percentage of typical cases are subjects positively identified, either with or without Rx records?
Overall, 65%...varying, of course, by the demographics of the applicant population.
Is there any value in knowing that a given applicant has been identified as being in the database but that he has no record of filling prescriptions?
Yes, more so than most underwriters realize.
No record in this context is usually good news unless, of course, the proposed insured has disclosed an impairment where one would expect Rx to have been dispensed. With this uncommon scenario, there are several possibilities:
- He may also procure Rx via health insurance coverage on his spouse (and that coverage may not be from a source accessible via Rx profile service).
- He may pay for the Rx out of pocket (which is quite expensive with the majority of drugs and hence distinctly uncommon when one is covered by a pharmacy benefit plan under health insurance!).
- He may rely solely on alternative and complementary (A&C) remedies, which are not prescribed and are purchased strictly over-the-counter. This may have profound underwriting implications and this underwriter implores insurers to include an application question about A&C indulgence!
- He may self-treat himself with non-A&C types of OTC medications and remedies.
- He may be given prescriptions and simply not fill them. In which case, if he says he is taking the drug and it does not come up on his profile, this may be sufficient reason to procure his APS!
Do Rx profile databases realize a high percentage of hits (individuals identified, with or without Rx records) in all age groups?
Hit rates are a tad lower at younger ages than in mid-life and later.
Recent Ingenix data show a stable hit rate between 60% and 70% for ages 40 and over. This, by the way, includes persons over age 65.
Do Rx profiles get higher yields in certain states?
A handful of states show yields under 50%. Conversely, there are states where 90% or more of applicants are found to have within the database. This may have implications for insurers who limit their business to a specific state or a few states.
If my company has an unexpectedly low yield of, say, 50% or less, doesnt this greatly diminish the value of Rx profiling?
Not at all..
What percentage of applicants has MIB codes? Does this influence your decision to use MIB? I hope not!
How many cocaine tests are positives? Does that low yield cause you to stop cocaine testing?
You see my point. The value of Rx profiling does not lie primarily in the % of applicants identified but in the quality and utility of the information on those who are identified!
Can I take adverse action on an Rx profile report?
No but, then, neither can you on an MIB code!
The Rx profile information is a clue. True, an extraordinarily credible clue but, still, a clue. It needs to be confirmed.
That said, confirmation may come in the form of denial. The applicant, determined to avoid disclosing information that might jeopardize his access to insurance, may firmly and repeatedly deny that the Rx was prescribed or, at least, prescribed for him.
This is why it is so valuable to have contact information on the physician who wrote the prescription.
Anecdotal experience suggests that underwriters may have great success using the telephone to contact the offices of attending physicians and confirm that a given prescription was written for the individual in question. In other situations, it may be highly desirable to secure the APS (and perhaps other requirements as well) when you sense that there is antiselection afoot, so to speak.
This is another way of properly amplifying the value of Rx profiling as it has come to be configured. You not only know which drugs you know who prescribed them!
Is there really any value in knowing drug dosage and how many times a day a drug is taken?
Emphatically, YES! This is one of the much-underappreciated benefits of Rx profiling.
Applicants often know what drug they take and how many times a day they take it but are stymied when asked what the dosage is. And that dosage can matter hugely.
Consider someone with a vague psychiatric history (perhaps the mother of all mental/nervous history vagaries: nervous breakdown) who has been prescribed one of the so-called novel (as in new and different) antipsychotics, such as olanzapine, risperidone and so on.
Does it matter if the patient is given a low dose vs. a high dose?
Well, consider that low doses are often prescribed off-label, in many cases for nonpsychotic conditions that may be readily insurable (at least for mortality risk). High doses, on the other hand, are reserved for patients with overt psychosis. Therefore, knowing that the proposed insured received a very low dose, especially for a limited duration, is a strong reason to pursue underwriting.
There are many other examples where both dose and times taken per day factor into risk assessment, so I for one am glad we have these bits of information in the mix of what is provided by Rx profiling. The knowledgeable underwriter with agree with me (as you will see from the report of the survey)
How common is it for a patient to be taking more than one drug?
Kaufman and his coworkers [Journal of the American Medical Association. 287,3(January 16,2002):327-42] did a telephone survey of a random sample of 2590 non-institutionalized American adults. They found that:
- 81% of their subjects had taken at least one medication in the previous week.
- Of that group, 50% took at least one prescription drug and 7% took 5 or more.
- Among women, 16% took 5 or more.
- At ages 45-64, 12% of women took 5 or more prescribed meds, and at ages 65+ 19% of men and 23% of women took 5 or more.
An interesting new study [Hajjar. Journal of the American Geriatric Society. 53,9(September, 2005):1518-23] looking at risk factors for frailty at older ages, found that the by far the best predictor from a list of 18 possibilities was having multiple Rx prescribers and that the number of drugs taken was also significant to the risk of premature frailty (the main underlying mechanism in geriatric mortality and morbidity)
Is knowing if the insured changed drugs for the same impairment of any value in underwriting?
Absolutely! This is one of the most important clues to impairment severity as well as whether or not control of the condition has been accomplished and sustained.
For example, lets say that a patient with asthma goes from using a commonly-prescribed adjuvant drug to taking oral steroids. This all but confirms a worsening of the asthma condition.
Another example would be a hyperlipidemic subject having a second lipid-lowering drug added to his existing use of a statin, or a person already using inderal for arrhythmia control being placed on a potent ventricular anti-arrhythmic.
Is it inherently more significant, in its own right, if an applicant takes a total of four or five drugs, as compared to none or just one, without regard to the reasons for which each is taken?
I believe the answer is YES, with two caveats:
- In most, but not all, scenarios this would apply only to drugs for chronic conditions rather an acute ailments
- This is more apt to be the case over age 50 or thereabouts, as compared to at younger ages
Why say this?
Because studies are now showing that there is a correlation between the extent of pain symptoms and morbidity, and even mortality.
The insured may have a variety of symptoms. They may each be tended to by a different drug. The fact that he takes four or more drugs related to chronic disorders associated with pain heightens my concern for insurability.
There have been studies that have shown how the number of medications taken as an independent factor, especially at older ages correlates with mortality.
Do I want to see all drugs prescribed to the applicant or is it better to have the unimportant drugs edited out of the profile to save underwriters some time when reviewing the Rx profile?
I am a underwriter. I want to see everything. I do not want anyone deciding, on my behalf, what is and is not important!
Lets say I have a 55 year old male who quit smoking 18 months ago and is eligible for non-smoker coverage. He takes a common aerosolized drug used for mild asthma. He says he has mild asthmatic bronchitis. Then I get the Rx profile and in addition to confirming the asthma Rx I also find a laundry list of clusters of antibiotics and related drugs used for protracted upper respiratory infections and pneumonia.
What do I now strongly suspect?
That the underlying impairment in this case is much more likely to be COPD than asthmatic bronchitis and I either do a belated drilldown interview of his history or I pursue his medical records.
Of course, I couldnt accomplish this if someone had decided for me that most classes of antibiotics were not relevant to underwriting!
Underwriters who covet edited Rx profiles concern me.
Managers who have never underwritten and mandate that Rx profiles be edited alarm me.
If the insured admits to one drug related to a given impairment, is there any additional value in knowing that related drugs are taken? In other words, isnt this overkill?
Contrary to mistaken notions held by some who manage underwriting but have never been underwriters, our credo goes something like this: the more I know about an applicant, the better I can determine his/her insurability.
Consider this example.
Jaber and his Mayo Clinic coworkers [Journal of the American College of Cardiology. 46,8(October 18, 2005):1473-78]report on the relationship between the medications given to patients who have had a coronary angioplasty and their prospects for being alive 36 months after the procedure.
Now, keep in mind that most applicants, even with abnormal findings on a PCI (percutaneous transluminal angioplasty) are nevertheless insurable on some basis not long after the procedure has been done.
What these cardiologists found was that there was a direct and highly significant relationship between whether or not these patients were prescribed the proper post-procedural drugs (and, if they were, how many of the four essential drugs were given) and patient mortality after 3 years.
Not to put too fine a point on the matter, it should be noted that those taking all 4 of the drugs had 33% lower 3-year mortality than those taking none or just one.
IV. The Implications of Nondisclosure
Having spent a good deal of time discussing this issue with veteran and savvy chief underwriters from North America and the UK, and having reflected on my own experiences and all that had been said, I am drawn to this conclusion: there are two kinds of non-disclosure and one is far more prevalent than the other.
The two kinds are inadvertent or, if you prefer, incidental nondisclosure (for which we have, as yet, no specific term) and willful or intentional nondisclosure (which we call antiselection). I believe that 90% of all nondisclosure in life and health insurance is of the first type.
And the mechanisms that cause this relate directly to how we have traditionally gathered risk information!
We ask questions of proposed insureds that they simply do not understand, we articulate these questions in ways that exacerbate their lack of understanding, we charge the wrong folks with asking them (wrong for several reasons, including skills and inherent self-interest), we provide insufficient space to record answers on paper applications and we often do not proactively go back and ask for clarification when we get partial or confusing responses.
We have several resources that work to counter incidental non-disclosure and one of the key tools is Rx profiling. I cite sufficient evidence to support this conclusion in this white paper.
The other form of nondisclosure is certainly more insidious (but not necessarily more damaging to the insurers bottom line), if only because it is premeditated.
Our resources to combat antiselection are really a precious few. The main ones are MIB codes, Rx profiles and well-designed teleinterviews.
Thirty-one (31) years ago, as a new underwriter, I had a case I will never forget. The proposed insured experienced a spasm of selective recall which catalyzed a vague statement on the application related to an otherwise undisclosed history of Stage IV Hodgkins disease (an impairment that was back then, not necessarily now, inexorably fatal and he knew it!).
He said he was taking citoxin, implying it was for flu-like symptoms (which some of the so-called Hodgkins B, or constitutional, symptoms certainly are).
There was no such drug in the Physicians Desk Reference (PDR).
I decided to try phonetic equivalents and finally stumbled upon Cytoxan, the proprietary name for cyclophosphamide, a major anticancer chemotherapy drug back in those days.
Further underwriting promptly disclosed the truth and the application was declined, thus avoiding the prospect of having to pay the claim if the insured survived two more years.
Today, I have a 70%+ chance of discovering this drug on an Rx profile. And once I look at the risk implications of cyclophosphamide, I am on my way to unmasking antiselection.
Rx profiling is a tremendous asset in terms of both prevalent (incidental) and uncommon (antiselection) forms of nondisclosure. In this way it is a major contributor to the underwriting process.
V. Off-Label/Unapproved Rx Prescribing
Extensive tracking of the medical literature and gradually developing a unique database on this subject, have come together to convince me that this is one of the gigantic Achilles heels in L+H underwriting.
"An estimated 23% if prescriptions are written for off-label uses."
The New Pill
U.S. News and World Report
April, 26, 2004; page 40
"The practice of clinical psychopharmacology is sufficiently ambiguous so that lots of patients can be prescribed new agents in good faith even when the evidence supporting the efficacy of those agents for those particular patients is still uncertain."
Lawrence H. Price, MD, Editor
The Brown University Psychopharmacology Update
A couple of examples may help to demonstrate the degree and significance of this reality of modern pharmacology.
According to the 2005 PDR, modafinil (Provigil ) is approved to improve wakefulness in patients with excessive sleepiness with narcolepsy, obstructive sleep apnea/hypopnea syndrome and shift work sleep disorder.
Therefore, an underwriter, learning of the prescribing of modafinil to a 35 year old person who is (1) of normal weight, (2) has no history of cardiovascular conditions associated with sleep apnea and (3) has an obvious daytime occupation, would pursue the likely prospect of this person having narcolepsy.
However, clinical studies have now shown that Modafinil is an effective adjuvant Rx in patients with daytime hypersomnolence caused by insomnia associated with major depression.
If this proposed insured also took a selective serotonin uptake inhibitor such as Prozac and attributed its use to transient difficulties associated with an ongoing divorce, the underwriter would be disinclined to link the two drugs and hence fail to further investigate the possibility of a potentially-serious impairment.
That is, unless the underwriter had a way to identify this likely-prevalent, but as yet unapproved, use of modafinil.
A much more common example of this phenomenon is the use of anticonvulsant (antiseizure, antiepilepsy) drugs. The number of impairments for which some of these drugs are used is a fraction of those for which they are currently approved.
Take for example gabapentin (Neurotin , which is also now available generically). This drug is often used in bipolar patients, as well as in the management of various settings of chronic pain such as chronic pelvic pain syndrome (prostatodynia in males, vulvodynia in females and potentially interstitial cystitis in either gender).
It is easy to see how an important mortality (bipolar) or morbidity (chronic pelvic pain syndrome) diagnosis could be missed if the underwriter was not aware of the full range of impairments for which gabapentin might be prescribed.
The Rx profile identifies the drugs that have been prescribed but it cannot account for off-label uses.
VI. Patient Adherence to Prescrived Drugs
"Drugs dont work in patients who don't take them."
C. Everett Koop, M. D.
Former US Surgeon General
It is obvious that for any drug to achieve its intended therapeutic effect, it must be taken, ideally as indicated in terms of daily consumption on the pill bottle.
This becomes very significant in contemporary underwriting because, more and more, we are making favorable underwriting decisions based in part on the presumption that a drug prescribed is a drug taken.
For example, with preferred risk underwriting paradigms, the vast majority of insurers (as confirmed in a recent survey undertaken by this author) will allow treated hypertensive and hyperlipidemic applicants to quality for preferred (or even super-preferred) coverage if they say they are taking certain antihypertensive or hypolipidemic drugs and appear to have adequate control of their condition (often based solely on a single current blood pressure reading or blood profile, respectively).
In consideration of the profound underwriting implications of Rx adherence, we will now examine the length and breadth of this issue in sufficient depth to make the point about how Rx profiling makes a major contribution by simply telling us if the proposed insured fills the prescription he says he takes, and, if so, if he refills it at intervals consistent with taking it as prescribed.
"Poor adherence to medication regimens is common, contributing to substantial worsening of disease, death and increased health care costs."
Adherence to Medication
Lars Osterberg, MD and Terrence Blaschke, MD
The New England Journal of Medicine
353,5(August 4, 2005):487-97
What are the costs of non-adherence to Rx?
Osterberg and Blaschke say that 33% to 69% of all medication-related hospitalizations are due to poor medication adherence, with a global cost to the health care system of $100,000,000,000.
Why are patients non-adherent?
"There is no one cause of poor adherence. Different causes are invariably operating in any group of persons."
ATP III Final Report National Cholesterol
Foundation Program Treatment Panel III
Some of the reasons cited again and again in the medical literature include forgetfulness, other priorities in life, intentional decision to limit doses, suspicion that symptoms concurrently experienced are due to the Rx, emotional factors and so on.
What are the various kinds of non-adherence?
- Not having the prescription filled
- Taking an incorrect dose
- Taking the medication at the wrong time
- Stopping too soon
Do we have any predictors of poor adherence?
Yes, Osterberg and Blaschke have identified some of the more prevalent predictors as follows:
- Depression or some other psychological problem
- Cognitive impairment
- Treatment of asymptomatic disease
- Inadequate MD follow-up with the patient
- Side effects of medication, real or suspected by the patient
- Patients lack of belief in the value of the Rx, often expressed to MD
- Patients lack of insight into his illness (which bodes ominously for disclosure!)
- Missed appointments
This underwriter would add: use of alternative and complementary therapies upon which the individual has come to have greater faith than he has in the conventional Rx prescribed by his physician.
The underwriting implications of some of these predictors are, obviously, highly significant to all mortality and morbidity risks.
How does coexisting depression (which may be undisclosed and difficult to pin down), influence mortality and morbidity in cardiovascular disease?
"Depression increases mortality through aspecific mechanisms such as non-compliance with medical treatment than through specific biological mechanisms affecting cardiovascular disease risk."
Does Depression Specifically Increase Cardiovascular Mortality?
David J. Vickers, MD et al.
Archives of Internal Medicine
165(January 10, 2005):110[letter]
Is non-adherence mainly a probably with less-educated people?
The above-cited National Cholesterol Foundation report observes that there are no significant differences in non-adherence based on gender, age, degree of wealth and thus, by inference, degree of education. Non-adherence is therefore a significant problem impacting all insured populations.
What % of patients are likely to be non-adherent in any cohort treated for the same reason with the same drugs?
The answer depends on the impairment and the nature of the Rx. Estimates range from < 5% to 75% or higher.
Is non-compliance more common in asymptomatic diseases, because the insured is less motivated to take the drug?
Some say YES, others say NO. Again, it depends on additional factors.
Is noncompliance higher in acute or chronic disease?
In acute conditions.
Chronic diseases usually mandate taking Rx for years, if not lifelong. Patients often become discouraged because of a perceived lack of efficacy of the drugs, and so on.
Is partial adherence to Rx use less significant than failure to ever fill the prescription?
Perhaps, but this citation certainly does not support the argument that partial adherence to Rx use is a minor matter:
"The effects of partial compliance are numerous and include lack of a physiologic response, increased risk of emergency room visits and hospitalizations, increased risk of the development of diseases and even increased risk of death."
Compliance with Treatment Regimens in Chronic Asymptomatic Diseases
Nancy Houston Miller, BSN
The American Journal of Medicine
102 (2A), (February 17, 1997):43-49
What do drug refilling rates tell us?
Osterberg and Blaschke state flatly that rates of refilling of prescriptions are an accurate measure of overall adherence...
Is denial of non-adherence credible?
The National Cholesterol Foundation report says directly that if a patient denies non-adherence, he/she is telling the truth about half the time.
For which impairments has non-adherence been shown to be significant?
I looked in the published literature on this subject and found that the impact of non-adherence is vast, impacting many of the most prevalent impairments we underwrite for mortality and morbidity.
Here is a list of a dozen major impairments (with one recent citation for each) showing what a major issue this is for underwriting:
- Asthma [Miller. Op. Cit.]
- Epilepsy [Ibid]
- Arthritis [Ibid]
- Lipids, where 30-40% or more are non-adherent [Jackevicius. Journal of the American Medical Association. 288,4(July 24/31,2002):462-467]
- Hypertension [Yiannnakopoulou. European Journal of Cardiovascular Prevention and Rehabilitation. 12,3(2005):243-49]
- Depression [Rosenow. Mayo Clinic Proceedings. 80,8(August, 2005):983-87]
- Heart Failure [Miller. Op. Cit.]
- Ulcerative colitis [Kane. The American Journal of Medicine. 114(2003):39-43
- Lupus Erythematosus [Shishov. Arthritis & Rheumatism. 52,9,Supplement (September, 2005):S188[abstract #423]
- Osteoporosis. [Silverman. Arthritis & Rheumatism. 52,9,Supplement (September, 2005):S298,abstract #737]
- Gout [Brewer. Arthritis & Rheumatism. 52,9,Supplement (September, 2005):S656,abstract #1771]
- Hepatitis C [Wiegand. Hepatology. 42,4,Supplement 1(2005):647A, abstract #1149]
How does noncompliance impact mortality and morbidity?
In a word: DRAMATICALLY!
Three major studies in cardiovascular disease underscore this important point:
In a study that followed hyperlipidemic patients taking clofibrate, those who were non-adherent to therapy had a 5-year mortality of 25% as compared to 15% in those who took the drug as prescribed [The Coronary Drug Research Group. The New England Journal of Medicine. 303(1980):1038-41]
In a later study involving beta-blocker Rx for MI survivors, those who did not adhere to the prescription had mortality that was 2.5 times that of drug-compliant patients. [Beta-Blocker Heart Attack Trial Research Group. Journal of the American Medical Association. 247(1982):1707-14]
A similar study in female MI patients showed that those who were poor adherers had nearly the same excess mortality (2.4 times greater than adherers) as seen in the immediate-previous 1982 investigation. [Gallagher. Journal of the American Medical Association. 270,6(August 11, 1993):743-44]
I am endeavoring to make these points about drug adherence as emphatically as I can because my experience in discussing Rx profiling as well as seeing the proprietary protective value studies done in-house by insurers show me, unequivocally, that this important aspect of the pay-back from Rx profiling is not at all well appreciated.
Hopefully, this will change with the circulation of this white paper!
VII. The X Factor
I have chosen this subtitle for Section VII because what I am about to discuss is still off the radar screens of most life and health insurers.
Which is unfortunate, given its importance.
In fact, I and a coauthor from a major actuarial consulting firm are hoping to change this misperception when we get to doing our proposed white paper on what I like to call alternative underwriting.
What is alternative underwriting?
It is the use of nontraditional risk-related information in the appraisal of insurability. By nontraditional I refer to factors that we have either not known to be relevant (pet ownership by older age persons) or which we have recognized as untenable (degree of physical activity) in an underwriting environment where 60-70% of risk information is gather by producers!
How does Rx use – and thus, Rx profiling – fit into this spectrum?
Here is a clear illustration of how important this topic is:
We have volumes of recently-published evidence that extol the so-called pleiotrophic effects of statins. These are the beneficial effects for the user than transcend the primary indication (hyperlipidemia) for statin use.
Indeed, these pleiotrophic effects are so numerous and so important than it is likely millions of patients will be taking statins on a prophylactic (preventive medicine) basis. (For more on this subject, see JournalScan, September, 2002, pages 2-4. )
Does the fact that one is taking a statin make him a better risk than is reflected in just the lipid-lowering effects? In my opinion, it sure does!
Back in 1986, three physicians, writing a letter to the editor of The New England Journal of Medicine, observed that their patients who took estrogen after menopause had a substantially lower risk of accidents, homicide and suicide. Clearly, as they say, there is no biologically feasible reason for a protective effect of this kind.
Nevertheless, their study findings said there was a reason and it wasnt the direct effect of hormone Rx per se. Rather, it was due to the fact that their willingness to take such therapy (then considered desirable and preventive of future problems) was a marker for other behaviors and modes of conduct that lessen the risk of trauma-related death.
Three Boston physicians believe statin use may confer a similar benefit beyond its known pleiotrophic effects:
"Long-term statin users tend to be healthier, less physically and cognitively frail, and more adherent to therapy and screening than nonusers."
Soko Setoguchi MD, MPH et al.
The New England Journal of Medicine
353,9(September 1, 2005):952[letter]
If, indeed, statin use is a marker for such considerations, then it does make a proposed insured an even better risk than just on the basis of strictly the direct medical effects of these drugs.
This idea is reinforced by McDermott, Schmitt and Wallner in their review paper on the Impact of Medication Nonadherence on Coronary Disease Outcomes.
"We suspect that patients adherent with medication regimens have other characteristics that confer a protective effect on survival.'"
Archives of Internal Medicine
The point here is that I believe, in time, considerations like consistently taking Rx when prescribed, having recommended screenings tests such as mammograms and other voluntary behaviors will be key elements in how we evaluate and assess risks for life and health insurance. When this happens, the value of Rx profiling will even be greater than it already is.
VIII. Rx Profiling and Teleunderwriting
Is teleunderwriting the way in which all insurers will gather, sort and assess risk information in life and health risk management?
Being described by some as a teleunderwriting guru, it goes without saying that I would say that this will be true, sooner rather than later!
I can support my enthusiasm here with two citations:
In the June, 2005 issue of LOMA Resource, Mary Holloway, a senior executive at Nationwide Life, reported that her company reduced its time service from 42 days to less than 20 days using a teleunderwriting process, which she describes in some detail.
In the September, 1998 issue of this same periodical, Jennifer Richards, Chief Underwriting Officer at Principal Financial, estimated that "…the information on the tele-application is 25 percent better than with the traditional [underwriting] process."
On the occasion of speaking at a subsequent seminar on this subject that I organized and convened in Chicago, Jennifer stated that after assessing the protective value of teleunderwriting, the conclusion was that, dollar for dollar spent, teleinterviews with drilldowns had more protective value than the MD reports they would have gotten had they not chosen to teleunderwrite!
The chief underwriter of a medium-sized and highly successful Southern company told me, in an email dated 9/1/05, that they have found Rx profiling has collaborated with teleunderwriting to produced what he calls significant improvements in our mortality. He goes on to say that "between teleunderwriting, [Rx profiling] and a lab the APS can be all but eliminated, adding that he would encourage all my counterparts to use this tool [referring to Rx profiling]."
One of the best things about teleunderwriting is that we are able to expand on the questions we ask proposed insureds once they have acknowledged any impairment. In the teleinterview drilldowns I have designed, great emphasis is placed on eliciting as much information on Rx both prescribed and OTC/A&C as possible. This information is collected, in part, with the specific intention of comparing it to what is provided on an Rx profile.
Taken together, the drilldown and the Rx profile greatly expand the basis on which we can triage the risk triage being a term I have tried to fit into the lexicon of teleunderwriting because it so aptly describes what we do after we have the screening requirements and have to decide whether to take action or pursue further elective requirements.
In a nutshell, the more times we take action instead of getting more information and overwhelmingly more information translates to a slow APS the more teleunderwriting contributes to our success.
Companies, not yet teleunderwriting, derive a clear and significant benefit from Rx profiling for all the reasons already cited. However, it is through teleunderwriting that Rx profilings impact on turnaround time and new business expense acquisition is experienced most profoundly.
IX. The Protective Value of Rx Profiling
In a recent effort to satisfy the needs of a client, I made an effort to determine the number of published protective value studies related specifically to underwriting requirements.
The list is shorter than I would have liked. 22 papers in just over 30 years and not one that addresses Rx profiling!
Please note that I said published studies.
I know of 2 large studies of Rx profiling that, to my knowledge, are unpublished. I have never seen either study, although I did make a direct request to review the data supporting the findings in the more recent one. That request was denied. Therefore, I cannot comment on either study.
There have been many informal cost/benefit studies done by insurers who have done pilots with one or more firms offering Rx profiling. I have personally seen several of these. My broad observation is that all of them that I have seen have been done incorrectly and as a result have significantly understated the value of Rx profiles.
Why was the value understated in these internal cost/benefit analyses?
For (at least) three reasons:
- The studies were too small not enough files reviewed
- Too much emphasis was placed on one aspect: detecting antiselection
- The persons who did the studies gave every indication of having, at best, a modest knowledge of pharmacology. They did not understand such key factors as the importance of dose, adherence, combinations of drugs, associations between drugs, and scenarios where off-label prescribing appeared to be present
There has been a protective value study on the MedPoint Rx profile. This was done by Reden & Anders, LTD, an actuarial consulting firm that, like MedPoint, is part of Ingenix. The study documents quite good value for small group health underwriting but has limited value in a wider context.
I have offered to conduct a basic underwriting study of this kind, focused on an underwriting analysis (rather than combining this analysis with actuarial input) but I have not as yet been engaged by anyone to do so.
What we really need is a thoroughgoing protective value study, merging the knowledge of a pharmacology-competent veteran underwriter and an actuary who can put credible numerical values on the impact of Rx profiles.
How should such a study be done without elaborate actuarial ROI analysis?
In my view, it would be done as follows:
- A significant batch of consecutive cases on which an Rx profile was obtained that contained Rx information would be collected.
- Rx profile information would then be compared to what else is known from all other information gathered to make the underwriting decision.
- This analysis would not be concerned with whether or not the underwriter of record actually used the Rx information (as it is inferred that many of the ways in which the profile would be an asset in risk analysis would not yet be appreciated by an underwriter).
- The analyst would estimate the magnitude of the impact in each instance where he believed the Rx profile added value, documenting the nature of that value and the degree it could have influenced the underwriting decision.
- The data from each case analysis would be collected, further analyzed and then reported in sufficient depth to establish whether or not the value of the Rx profile was significant to justify its deployment in the companys underwriting arsenal.
This is not, as one can see, a very complicated undertaking. That is, if you are a very sage underwriter with a solid foundation in pharmacology from a risk appraisal perspective
X. Recommendations as to How to Interface Rx Profiling with Your Underwriting Process
If and when you decide to examine and potentially embrace Rx profiling, these are some of the things I think you need to do to maximize its impact on your bottom line:
- Do an Rx profile trial and conduct a sufficient cost/benefit study to accurately determine its value to your company in your market and underwriting environment.
- If the trial proves successful and the value is appreciated, go on to step 4.
- If it does not show sufficient value, get a second opinion before you accept the findings (even if it is done by a medical director).
- Upon proceeding with Rx profiling, integrate it fluidly into your underwriting workflow system.
- While you are doing that, look at your options for providing some type of Rx-related education for all who will be involved in the use of this information. This, of course, means all underwriters as well as internal teleinterview staff, all individuals who make any jet or simplified issue decisions, or who have higher-level field contact on behalf of underwriters (underwriting assistants and so on).
- Create definitive guidelines so that underwriters appreciate all aspects where Rx profiles impact cases; conduct hypothetical case clinics before underwriters actually use Rx profiles so that their comfort zone is adequate.
- Audit the results of Rx profiling very early on. Certainly some sort of value study should be done in the first six months (unless, of course, a very favorable study came out of the trial period; then it may be acceptable to wait up to a year before doing a current study).
- Get feedback from underwriters on an ongoing basis. Make this a definite part of period meetings with underwriters. Ask them to present cases where the Rx profile had an impact, so that all underwriters will have an appreciation of how to get maximum value from the Rx records.
- Consider making use of Hanks Underwriting Guide to Rx when it is available, as it will address some aspects of Rx profiling that need to be covered and are seriously loose ends at the moment (most notably, off-label use of Rx).
- If all goes well, make Rx profiling an intrinsic part of your screening of some or, ideally, all underwritten risks. This includes jet issue cases, simplified issue products, etc.
XI. Report of a 52 Company Rx Profiling Survey
No white paper looking at the merits of a relatively novel underwriting resource would be complete without feedback from the underwriting community.
Accordingly, I and my colleague Esther undertook a rather extensive survey of a cross-section of life and health companies to gauge their impressions about the pros and cons of Rx profiling as well as their experience if they had done a trial of such a service or were currently using Rx profiling.
Fifty-two (52) life and health insurers voluntarily participated in the survey.
Please note that some respondents did not answer all questions and others gave more than one answer to some questions. This is why, if you tally up the responses to any given question, the number might not add up to 52. We check and then double-checked our tabulations and we are confident we got them right.
This section shows the actual questions asked of survey participants as well as their answers, followed by summarized comments received from respondents.
Comments by this underwriter are in bold italics.
PART 1: Perceptions
Please indicate your response to the following 12 statements about Rx profiling as a potential case underwriting resource:
1) From what I know and/or have heard, Rx profiling sounds like it could be a cost-effective underwriting resource.
- 12 strongly agree
- 27 agree
- 6 no opinion
- 6 disagree
- 0 strongly disagree
2) Rx profiling would be especially effective in a teleunderwriting context.
- 17 strongly agree
- 25 agree
- 5 no opinion
- 4 disagree
- 0 strongly disagree
It is reassuring to see that when we ask about Rx profiling in the specific context of teleunderwriting, the number of strong agreements increases from 12 to 17!
3) Rx use details on proposed insureds from all sources represent independently valuable case underwriting information.
- 24 strongly agree
- 21 agree
- 4 no opinion
- 2 disagree
- 0 strongly disagree
This is even more satisfying as I have been speaking and writing on this subject for many years.
4) Rx information provided via non-medical, paramedical and medical underwriting is often inaccurate and/or incomplete.
- 5 strongly agree
- 22 agree
- 8 no opinion
- 12 disagree
- 0 strongly disagree
While the consensus is, as one would have to expect, that Rx information from traditional sources is replete with inaccuracies and is often significantly incomplete, it is a bit disconcerting to see that 25% of respondents do not recognize this reality.
5) The fact that there is little accessible information on off-label (unapproved) uses of medications makes it more difficult for underwriters to relate some prescribed drugs to the impairments for which they are being used.
- 9 strongly agree
- 26 agree
- 10 no opinion
- 5 disagree
- 0 strongly disagree
Seventy-five percent (75%) of respondents appreciate that a lack of knowledge about off-label uses of drugs is a major factor in limiting the underwriters ability to make effective use of Rx information. I would think that those who have no opinion may not have an appreciation of the prevalence of off-labeling prescribing (which we have already addressed in this white paper).
6) It would be helpful to have a resource that fully explained the relationship between medications and their potential effects on lab test results.
- 19 strongly agree
- 27 agree
- 3 no opinion
- 0 disagree
- 0 strongly disagree
7) One reason we are concerned about using Rx profiling is that we will experience pushback from insurance applicants.
- 1 strongly agree
- 8 agree
- 12 no opinion
- 25 disagree
- 4 strongly disagree
This was one of the most important questions in this section of the survey, because this was one of the premier concerns of insurers when Rx profiling first became available. It is reassuring to know that over 80% of respondents either agree this is a non-issue or havent formed an opinion to the contrary.
8) A major disadvantage with Rx profiles is that providers do not get hits (defined as finding the proposed insured in their data base, whether or not Rx has been prescribed) on a significant % of proposed insureds.
- 13 strongly agree
- 15 agree
- 13 no opinion
- 8 disagree
- 3 strongly disagree
Perhaps I did not have to make this point as emphatically as I did earlier in this report. Most companies appear to see that it is not the number of hits but the content of those hits that drives the value of Rx profiling.
9) All the redundant reports of medication refills on Rx profile reports have no value in underwriting and simply make Rx profiles take longer for underwriters to review.
- 0 strongly agree
- 7 agree
- 12 no opinion
- 24 disagree
- 8 strongly disagree
While it is good that 60% appreciate there is value in Rx refill information, more emphasis on this aspect of Rx profiling needs to be made so that everyone sees the hidden (to some, anyway) pay-off here.
10) We are concerned about the risk of Rx profiling eventually being made a public issue by journalists, perhaps even on 60 MINUTES!
- 3 strongly agree
- 12 agree
- 19 no opinion
- 18 disagree
- 4 strongly disagree
If Rx profiling were new, this would be a very valid concern.
However, given that it has been around now for years and moreover that there have been very few anecdotal reports that have come to my attention regarding applicant concern when being told of this process in conjunction with their assessment as risks, the likelihood of something like this happening gets less and less with each passing day.
11) All the details of dosing how many times a day, size of the dose (as in 10 mg vs. 40 mgs), whether timed-release or not are of no real underwriting value and simply make Rx profile reports take longer for underwriters to review
- 2 strongly agree
- 8 agree
- 3 no opinion
- 26 disagree
- 12 strongly disagree
At the risk of being redundant Bravo one more time!
12) Our competitors dont do it and therefore we will have problems with our producers if we do Rx profiling
- 2 strongly agree
- 6 agree
- 8 no opinion
- 27 disagree
- 8 strongly disagree
There has always been and still is a strong emphasis placed on both what producers will think as well as what key competitors do when it comes to which requirements insurers insist upon..
Fortunately, almost 70% of respondents would not allow themselves to be unduly influenced these two factors when it comes to using Rx profiling.
Further, as regards agents and brokers, all of my experience which includes speaking three times at the Million Dollar Round Table and countless hundreds of times to other producer audiences would argue that this is not the type of requirement that would raise their ire. This is because (1) it does not impose anything that must be done on their part or that of their clients, and (2) it looks at highly-objective information that merely confirms and expands upon questions already asked on almost every application, at least for fully-underwritten business.
PART 2: Experience with Rx Profiling
Please indicate which of these responses best fits your companys situation at this time:
- 17 - We presently make use of Rx profiling in case underwriting on some basis If this is your response, please complete Section A below
- 2 - We used to do Rx profiling in case underwriting on a regular basis (not just for the purpose of doing a trial to evaluate the service) but we stopped doing it If this is your response, please complete Section A below
- 1 - We have done one or more trials of Rx profiling services and are considering adopting this resource for case underwriting purposes If this is your response, please complete Section B below
- 4 - We have done one or more trials of Rx profiling and decided not to adopt this resource for case underwriting purposes If this is your response, please complete Section C below
- 27 - We have never done a trial of Rx profiling If this is your response, please complete Section D below
NOTE: WHEN YOU COMPLETE YOUR DESIGNATED SECTION BELOW, YOU HAVE COMPLETED THIS SURVEY.
In the interest of keeping this white paper in a manageable length, we will not review the solitary response to the third question. The answers of the 2 YES respondents to question two are melded with the YES respondents to question one. We will look at why 4 companies had a trial and decided not to use Rx profiling and also at the responses from those 27 carriers who have not yet done a trial run with this service.
The comments given by respondents will be reviewed after we look at the replies in Sections A, C and D.
Which best describes your company situation with regard to Rx profiling?
- 15 - We do it currently
- 3 - We used to do it but we stopped
1. How long have you been doing or did you do Rx profiling in case underwriting?
- 5 - < 6 months
- 2 - 6 months 1 year
- 7 - 12 years
- 4 - 24 years
- 0 - > 4 years
2. Do you/did you do it on all 0 - or just some 17 - cases?
I am not convinced that we all know when to use Rx profiles; in fact, I cannot envision a scenario in underwritten business where Rx profiling is inherently contraindicated.
3. If you do/did it on just some cases, which of the following best describes the circumstances where you use/used Rx profiling?
- 1 - On one product only
- 17 - At a minimum or maximum face amount threshold
- 1 - Within a certain applicant age-range only
- 0 - Only if medications are/were admitted to by applicants
- 5 - Only if medications are/were not admitted to by applicants
- 2 - On teleinterviewed cases only
- 0 - On non-teleunderwritten cases only
- 11 - Other
The perception that requirements should be used at some face amount threshold is both ingrained and logical with the caveat that we have yet to test the premise that Rx profiling produces significant-enough value to be justified as a universal requirement, akin to medical history-taking or MIB code use.
4. Have you done any internal cost/benefit or protective value studies on Rx profiling as a case underwriting resource?
- 5 - YES
- 14 - NO
One would expect this answer. However, if I were a chief underwriter and began to use any new requirement, I would see great merit in proactively scrutinizing its value early on.
5. If YES, have you found this resource to be cost-effective?
- 1 - YES, significantly
- 3 - YES, marginally
- 2 - NO
I think you can chalk these responses up to the fact that most companies do not have a broad enough appreciation of the ways in which Rx profiles contribute to protective value and hence they tend to undervalue them.
6. How do/did your underwriters like this resource?
- 13 - They like it
- If they like it, what would you say is/was the #1 reason?
Note: All subjective question comments will be considered following this report of objective responses to survey questions
- 7 - They dislike it
No doubt a large part of underwriter dislike is rooted in the fact that there is often considerable information provided on some proposed insureds and also that underwriters lack the resources to adequately assess the implications inherent in such profiles. Then, too, there is the very real fact that demands of production may make cause any additional risk information, however desirable, to be perceived as a hindrance to getting out the work!
- If they dislike it, what would you say is/was the #1 reason?
7. Have you/did you encounter(ed) any problems with Rx profiling?
- 6 - NO
- 12 - YES
If YES, please briefly describe this/these problem(s)
8. All things considered, would you recommend Rx profiling to others?
- 12 - YES
- 4 - NO
Note: 2 companies provided a third answer: depends
This response speaks for itself. At the end of the day, at least 75% of respondents found Rx profiling of such value as to make it worthy of recommendation to peers.
1. Are you/did you run your Rx profile trial in a teleinterview environment?
- 3 - YES
- 1 - NO
2. Roughly how many cases did you use/are you using in your trial?
- 0 - < 100 cases
- 2 - 100 200 cases
- 0 - 200 500 cases
- 2 - > 500 cases
3. How long did your trial run or has it run so far?
- 1 - < One month
- 2 - 13 months
- 1 - 36 months
- 0 - > 6 months
4. How do your underwriters like this resource?
- 1 - They like it
If they like it, what would you say is the #1 reason?
- 3- They dislike it
If they dislike it, what would you say is the #1 reason?
5. What is the #1 reason you decided not to make use of Rx profiling after completing your trial?
- 0 - Too few hits on our applicants
- 0 - Too little worthwhile underwriting information
- 1 - Underwriters were unable to effectively use this information
- 0 - Appears to be too expensive (but we did not do an internal cost/benefit analysis)
- 2 - Not cost-effective based on our internal cost/benefit analysis
- 0 - Other
Certainly the issues identified here can be remedied easily enough.
6. Do you have any other comments you would like to share on this subject?
1. Have you ever been approached by an Rx profile provider to conduct a trial?
- 15 - YES
2. If YES, how long ago were you approached?
- 7 - Within 6 months
- 2 - 6 months 1 year
- 7 -> 1 year ago
3. Would you be inclined to undertake such a trial if asked?
- 7 - YES
- 12 - NO
Note: 4 companies said maybe
4. If NO, why would you not be inclined to undertake an Rx profile trial?
- 2 - Heard from others that it is not worth doing
- 3 - Price quoted for using Rx profiling is too high
- 0 - Concern for negative agents reaction if we do Rx profiling
- 2 - Concern for negative customer reaction if we do Rx profiling
- 1 - Concern for adverse regulatory reaction if we use Rx profiling
- 6 - Do not believe this additional Rx information adds that much value to that we normally get from applicants regarding their use of Rx
- 8 Other
The fact that 6 companies do not believe that Rx profiling adds value in addition to what is learned about Rx use from conventional sources is perhaps the single biggest argument why every company should at least do a trial of Rx profiling before making subjective assumptions.
There were a ton of individual comments. Many were redundant and some were vague or irrelevant. I will present only a portion of the more illuminating comments here, with my own comments added.
Note: The complete tabulated results containing all the comments are on file at my office and will be faxed to anyone representing a direct-writing life or health insurer who wishes to see all of the comments. Just send your request to Esther and provide her with your fax number.
1. One of the consistent responses under what circumstances best describe your use of Rx profiling was that they did it electively, at underwriter discretion or when there were RED FLAGS on the case.
I can appreciate why this would be appealing. After all, we do elective ordering of every requirement we have except the baseline risk question history and MIB codes. That said, I do not believe that underwriters have a basis for discerning when an Rx profile is apt to provide value, at least not in most cases. Therefore, I would strongly advocate that the ideal way to use this tool is as a screening test akin to blood profiles or motor vehicle records, not as an elective or, if you will, reflexive resource.
2. The primary comments on why underwriters like Rx profiling were that profiles disclosed additional salient risk information
What can I say here but amen?
3. The sole reason cited in comments about why underwriters did not like Rx profiling had to do with perceived low hit rates.
There was also the matter of cases where the insurer was told the applicant was on Rx and that medication did not come up on the profile.
We addressed the first matter earlier.. Said succinctly, it is not the number of hits but the content of the hits that matters, just like MIB codes or MVRs
As regards the question of medications having been given but not reported on the profile, I indicated earlier on in this paper the primary reasons why this happens. No doubt the very same thing happens with MVRs, albeit for different reasons.
Certainly if one is told by the insured that he takes an Rx which is not shown on the Rx profile, one would be ill-advised to not accept what has been acknowledged, just as when one says they smoke and then they test negative for cotinine, we do not issue at non-smoker/user rates!
4. With regard to problems encountered with Rx profiling, there were three basic negatives aired:
- Low hit rate
- Technical problems, that should be resolvable
- Issues related to family plan Rx benefit coverage where the reports contained information about other family members not applicable to the insured (a nagging concern that cannot always be logically sorted out)
5. On the matter of recommending Rx profiling to others, one company observed that hit rates vary by product and market, and that, as an example, they were much higher on DI applicants than on life cases. One company gave an enthusiastic thumbs up while another said its recommendation would be qualified dont expect too much.
There is no doubt that some companies, and some products as well will have higher hit rates than others but, at the risk of redundancy, I implore everyone to value the hits individually, not the % of hits unless it so very low, say under 20% or even 10% where the costs of maintaining the requirement as a resource are not justified. One company said it almost never got a hit and that is certainly due to some company-specific extenuating circumstance.
6. From those who responded to why they chose not to embrace Rx profiling after a trial period of using it, the main comments had to do with low hit rates and lack of perceived cost-effectiveness (usually subjectively, one might add).
If you do a trial of any requirement, you do it long enough to get credible results and then undertake some sort of informed analysis of its value (or lack thereof). The chief problem as, as explained in depth earlier, that most companies doing such analyses do not appear to have the needed insights into the ways in which Rx profiling contributes to risk assessment. In that absence of such insights, the focus always ends up being unduly on detecting perceived significant undisclosed drugs and unmasking apparent antiselection. This, alas, is far too narrow a perspective to do a credible study of the value of Rx profiling.
7. On the matter of being disinclined to undertake a trial of Rx profiling, the responses centered on lack of insight into the potential value, fear of getting low yields, and concern for their underwriters ability to get value out of the product.
There was not one credible reason presented for not undertaking a trial; having said this I am empathetic to the reasons respondents believed that a trial would be a waste of time.
Rx profiling is like any other new requirement. There is a learning curve. This is why we rigorously consider every pro and con before we proceed; then, if the pros outnumber the cons, we do a meaningful participative study (in this case, a trial) and make our decision after we have adequately analyzed the results.
XII. Closing Comments
This is a time of great challenge to those entrusted with managing the profitability of health insurance. We are charged with carefully selecting risks while also keeping business acquisition costs with tight boundaries, expediting cycle time from application to issue of coverage and avoiding wastage through maximizing our insights in to the nature of each risk.
It is my hope that this in-depth look at Rx profiling will energize readers to undertake assessments of this underwriting resource in their companies. My research convinces me that Rx profiling ideally teamed with teleunderwriting will allow us satisfy all of these accountabilities and take what we call underwriting to a new level of efficiency and effectiveness.
I welcome questions and comments from readers.