Quantitative prognostics: an analytic defence


Barry Fogel

Quantitative prognostics: an analytic defence

PointRight examines how quantitative prognostics can protect nursing homes from the loss costs of professional liability.

Professional liability (PL) losses continue to be a significant cost item for skilled nursing facilities (SNFs) in many US states. In some states, tort reform has had exceptions for negligent nursing home care, whilst in others, the caps on non-economic damages still allow room for substantial verdicts. For example, damages in wrongful death cases may be multiplied by the number of surviving relatives. Overall, PointRight has found that SNFs’ greatest challenge with PL losses is not mega-claims of $10 million or more, or even claims of more than $1 million, but rather a large number of claims that close with indemnities of between $250,000 and $1 million. The causes of those claims are almost always some combination of pressure ulcers (bedsores), falls with serious injury, or death. The typical plaintiff’s claim is that these outcomes would not have occurred but for the facility’s negligence, with this claim of negligence often resting on a lack of documentation of preventive actions.

In PointRight’s experience, PL suits against nursing homes usually have some merit, but in some cases, the nursing home simply was not responsible for the outcome cited in the claim. In other cases, the nursing home made a minimal contribution to a chain of events in which other providers—hospitals and doctors—had a much greater role. Obviously, when either of these propositions can be proved by solid evidence, the value of the case will be much lower and the likelihood of a mega-verdict will be greatly reduced. When the plaintiff’s attorney is confronted with this evidence, a settlement for less than the case reserve is the usual outcome.

In some cases, a ‘smoking gun’ of physician error or hospital-acquired complications will be found on record review, but in most cases, the result of an expert review is at best an opinion that mitigates the SNF’s responsibility—an opinion subject to challenge by the plaintiff’s equally qualified expert. The key to breaking the tie between conflicting expert opinions is the introduction of quantitative analysis of the resident’s prognosis with typical, non-negligent care. Establishing a poor prognosis removes the presumption that if there was a bad outcome, the SNF must have done something wrong. With the presumption of liability removed, incomplete documentation—which is in fact the standard of practice and not the exception—can no longer be used to imply there must have been a failure to provide necessary services.

Quantitative analysis of prognosis—or quantitative prognostics— means replacing opinion with numbers. For example: “more than 50 percent of patients with this combination of diseases, conditions and functional impairments will be dead within six months” or “over the course of a year, more than 50 percent of patients, who, like the plaintiff, have morbid obesity, diabetes, hypertension, end-stage renal disease on dialysis, incontinence, and the need for two people to help with bed mobility and transfers, will develop a pressure ulcer. This incidence rate applies equally to nursing homes with one-star ratings and those with five-star ratings.”

These quantitative estimates are based on predictive models that use variables from the minimum data set (MDS), the mandatory standardised computerised assessment of health status that must be performed periodically on SNF patients as a condition of the facility receiving Medicare or Medicaid reimbursement. The MDS is part of the patient’s official medical record. While it is almost always reviewed by expert witnesses, it is seldom used as a basis for quantitative prognoses using predictive models. Nonetheless, such models are easily built if one has a sufficiently large MDS database, such as PointRight’s database of more than 18 million MDS records from more than 3 million discrete patients.

Introducing a quantitative mortality prognosis can help in four ways. First, it challenges the assumption that the resident’s death was the facility’s fault, rather than nature taking its course. Second, it makes gaps in documentation less damning. Third, it implies that the damages for wrongful death if the facility was liable are not great, because the patient did not have long to live in any case. Fourth, when a patient living in the facility ‘beats the odds’ by outliving their prognosis for several years, it is more difficult to claim that the patient’s eventual death proves the facility systematically neglected the patient. Combining these points narrows the focus of the case, limits the facility’s responsibility and reduces the perceived damages.

A quantitative prognosis for pressure ulcer development works similarly. If more than half of the patients with a given clinical profile will develop pressure ulcers over a one-year period, it is hard to argue that a facility is to blame for the patient’s ‘skin failure’. While the plaintiff may call attention to some intervention that was not made— or at least not documented—the burden is on the plaintiff to show that the pressure ulcer would not have developed if the extra treatment, a special pressure-relieving mattress for example, was provided. Such proof is difficult; if it were easy, the manufacturers of those special,expensive mattresses would be quick to provide it. If plaintiffs claim that extraordinary care and good luck is the standard of care, common sense works against them. Ironically, it is understood that modern medicine does not have the cures for all diseases or the solutions to all clinical problems, yet more is expected of SNFs than of hospitals even though SNFs have far fewer resources. In cases where the quantitative prognosis for pressure ulcers is poor, it’s worth checking hospital records to see if ulcers developed or worsened while the patient was in the hospital. Such an observation supports the idea that a pressure ulcer could develop or even progress despite an SNF providing an adequate level of care.

Quantitative prognostics can be used the same way to defend claims involving falls, but they can also be used in a different and equally potent way. It sometimes happens that a patient’s risk of falls is moderate—not so high as to warrant intensive fall precautions or special treatment, but high enough that over a period of months or years, a fall is likely to occur—leading to an injury in some unfortunate cases. Establishing that the risk is moderate implies that the occurrence of a specific fall is not evidence of negligence; establishing that it is not high can justify a facility not making a resource-intensive and potentially intrusive intervention in an attempt to prevent falls. It would not make sense to continue such an intervention over years when a patient was at moderate risk. If during several years of moderate risk, the patient did not fall, the facility should in fact receive credit, even if there was an eventual fall. Another quantitative point can be raised—the low proportion of falls that lead to serious injury. Total prevention of falls would require a complete loss of privacy, independence, and dignity, and would be prohibitively expensive. Given the low proportion of falls that lead to serious injury, it would not be warranted even if it were feasible.

Juries or families of newly admitted SNF patients may begin with a naïve belief that being in an SNF means they no longer need to worry about falls. In fact, the risk of falls in the facility will usually be lower than in the community, and the probability of injury will be lower as well, but falls nevertheless still occur. And, by the way, restraints don’t prevent falls, they just make them more dangerous if they happen. This point, too, can be supported by the numbers and by published geriatric literature.

The value of a prognosis-based defence will naturally vary by venue, by the extent of actual negligence, and by the skill of the plaintiff’s attorney and their familiarity with analytics-based defences. In addition, the state’s rules concerning contributory negligence and joint and several liability will be relevant in most cases. Notwithstanding, an analytics-based defence, emphasising a quantitative prognosis can be a game-changer in enough cases to significantly impact a captive’s loss ratio.

Because analytics-based defences using the MDS data are so often helpful, it makes sense for the captive to have members’ MDS data on hand at the outset and for MDS data to be collected by the facilities with adequate quality controls. PointRight’s experience supports the value of a captive-wide mandate for member facilities to utilise a prospective MDS data integrity auditing system, and to submit their clean MDS data to a consulting analytics organisation for use if, and when, needed to defend a claim.

For the past 16 years, PointRight has been serving the nursing home industry with data analytics—beginning with periodic measurement and benchmarking of facilities’ clinical performance and moving on to web-based data quality auditing, financial analysis, and predictive analytics for clinical risk identification and care planning.

Barry Fogel is executive vice president and founder at PointRight. He can be contacted at: barry.fogel@pointright.com

PointRight, professional liability, captive, insurance, Bermuda

Captive International