Underwriting mental health
The underwriting of mental health conditions has always been notoriously difficult. To properly assess such applicants an underwriter needs to look at the complete picture of not only the applicant’s health but also factors such as employment record and support network.
But in these days of increasing automation, insurers are under increasing pressure to accept as many cases as possible through an automated process, driving up their straight-through processing (STP) rates and keeping costs down.
An engine-based process involves only a few drill-down questions. But mental health risks tend not to lend themselves to this approach very well, if at all. Each applicant is different, their story different, and to make a decision based on four or five questions really isn’t satisfactory, let alone fair. We hear that the practice of some insurers amounts to identifying the really good risks and accepting at standard rates, and declining most of the rest. After all, referring all of the cases would have a significant impact on the STP rate, by which statistic success may be measured; remember that an auto-declined case counts as one that has been ‘straight-through processed’.
This way of doing things is, at best, poor for applicants and is as likely poor for insurers and the industry as a whole. Let us be clear: this is not to query principle of automation and its ability to deliver rapid acceptances and a painless customer journey. But for some individuals, with their inevitably individual health histories, it doesn’t make sense. And accepting mental health risks at standard or decline with nothing much in between would not stand much scrutiny by regulators.
These issues have been brought into sharp focus by a number of developments. First, as part of the recent Royal Commission in Australia, submissions were sought from interested parties. Those responding included Mental Health Australia and the mental health pressure group Beyond Blue, as well as individuals. Some of the submissions were very critical indeed. Second, a session at a recent conference in London dealt with underwriting of mental health, prompted partly by several press articles questioning the fairness of underwriting practices.
Given the prevalence of mental disorders in developed societies – for example, in the UK it is estimated that every week as many as one in six adults experiences some sort of common mental health problem – it is maybe no surprise that calls for fairer underwriting handling have finally become so vocal. The demand is that each case should be treated on its merits by consideration of all the relevant facts, rather than a ‘one size fits all’ approach.
All this interest from outside the industry raises the question of whether underwriting guidelines are underpinned by a robust, up-to-date evidence base. Are the ratings in underwriting manuals today based more on received underwriting wisdom than on recent data? We suspect so, although in fairness to the compilers of manuals, good mortality studies are hard to come by and morbidity ones as rare as hens’ teeth. While it may be possible to gauge the level of risk for a group overall, the lack of anything approaching granularity in the available data means that finer risk stratification is difficult indeed.
So what is the answer?
Underwriting cases on three or four questions via an engine appears not to be right. It is inconsistent with treating customers fairly and accurate risk selection. And asking questions like “Have you tried to harm or kill yourself?” is maybe heavy-handed and insensitive regardless of the underlying diagnosis. Perhaps this is where tele-interviewing really has its place, with a trained interviewer, especially a nurse, talking to the applicant and sensitively teasing out information and recording the story.
In underwriting manuals rating guidelines need to be reviewed and better researched to reflect up-to-date evidence and treatments. They need to be presented in such a way that they avoid a binary ‘black/white’ approach and assist the underwriter to determine the shades of grey.
We intend to keep an eye on developments here and will also be undertaking some research of our own. We will report back in due course, but in the meantime we would be interested in hearing your own views on this high-profile topic.