Doctors Are Not Experts on Life Expectancy

by David Strauss, PhD, FASA and Robert Shavelle, PhD

This article was originally published in the summer 1998 issue of the Expert Witness.

In personal injury cases, courts have traditionally relied on doctors for opinions on plaintiffs’ remaining life expectancy. We show here that such questions are really beyond the expertise of physicians, and that their testimony is readily challenged.

The analogy with life insurance is helpful. When applying for a life insurance policy you are first examined by a doctor, who assesses various risk factors. The results are transferred to the insurance company’s actuaries, who use the risk profile to assess your survival prospects. Thus both medical and statistical/actuarial skills are needed. Only a physician is qualified to appraise the individual, and only a statistician or actuary is qualified to turn the appraisal into a life expectancy.

Most physicians readily agree that they are not expert in actuarial issues. Nevertheless, pediatricians are still routinely asked to testify on the life expectancy of children with birth defects, while therapists or other medical specialists are consulted regarding adult accident victims. Their testimony on what are really statistical issues is often unfortunate. The following examples, with some modification, are drawn from actual cases.

“As a gerontologist I work with elderly persons. All the persons with cerebral palsy that I examine are at least fifty years old. Therefore I believe that this child with cerebral palsy will probably live to at least 50.”

We pass over this in silence.

“I believe that this child will certainly live to age 40, although probably not to age 50.”

It is, of course, absurd to say that any child – even one in perfect health – will “certainly” live to any age. Further, the probability that the age at death will fall in a narrow range such as 40-50 is bound to be quite low. The statement seems to confuse the life expectancy, which can often be estimated with some precision, and the actual age at death. The latter can rarely be predicted with any accuracy.

The annual mortality rate for children like the plaintiff is 1%. After 50 years, therefore 50% [ = 50 x 1%] of such children would have died. The median survival time is thus 50 additional years.”

There are two mistakes here. First, the math is wrong: in fact, 99% of the current survivors will survive one additional year, and therefore the proportion surviving 50 years is 61% (=.9950), not 50%. Second, the analysis ignores the dramatic increase in human mortality with age. As a result it gives wildly unrealistic long-term estimates, predicting, for example, that 37% of the population will survive to age 100.

* * *

Witnesses lacking statistical or actuarial training are frequently unable to define life expectancy, compute it in a simple case, or distinguish it from the median survival time. This may be exposed with a simple illustration.* If the witness cannot even explain what a life expectancy is, the testimony will lack credibility.

A physician’s opinion will be based either on a reading of the research literature or “on the basis of my clinical experience.” In the former approach, the plaintiff is matched to some group of individuals whose survival has been studied and reported. There are, for example, several studies of long-term survival for persons with cerebral palsy, traumatic brain injury, and spinal cord injury. Unfortunately such studies provide at best a crude estimate of life expectancy. The attorney can establish that:

  • The studies generally follow a cohort of persons who initially were of a given age and in a given condition. If the plaintiff is older and currently in this condition, it would be necessary to assume that cohort members surviving to the plaintiff’s age are still in that same condition. This assumption may be quite unreasonable, especially for young children who may have fair prospects for improvement.
  • Most studies provide survival curves, giving the fraction of persons in the cohort who survive to a given age. This will provide a median survival time only if the mortality is so high that 50% of the subjects die within the study period, and it rarely will permit the computation of a life expectancy.
  • The cohorts studied in literature are necessarily based on coarse classifications of one or two risk factors. Ironically, the clinician’s strength – the ability to make fine judgements about numerous patient characteristics – does not come into play.

It must therefore be recognized that published articles provide at best a rough approximation to a given plaintiff’s life expectancy. Indeed, some of the articles include a warning to this effect, a point that the opposing attorney may wish to emphasize.

Clinicians who instead rely on their experience for opinions are even more vulnerable. The lack of a solid basis can be revealed with questions such as:

  • How many patients closely resembling the plaintiff have you examined? [The answer will be at most a few dozen.]
  • Did you follow up on the survival or death of all of these patients? Give the specifics of your procedure. In particular, how did you follow the patients who moved to a different town or even to a different state? How did you ascertain who died? Where and in what form did you keep your records of the children’s survival time? Did you periodically reassess their functional levels during the follow up?
  • If you have been practicing for 20 years (say), how could you have ever observed a child surviving more than an additional 20 years? Does this lack affect your opinion? Why or why not?
  • Are you aware of the literature on statistical methods for estimating survival probabilities? Which methods did you use?

Such questions should make the limitations of the doctor’s expertise very clear.


* As an example, if 1/3 of members of a population will live exactly 2 more years, 1/3 will live exactly 3 more years, and 1/3 will live exactly 10 more years, then the life expectancy is (2 + 3 + 10)/3 = 5 years and the median is 3 years (the middle value). [Back to text]


1. Hutton JL, Cooke T, Pharoah POD. Life expectancy in children with cerebral palsy. British Medical Journal 1994; 309:431-435.

2. Chrichton JU, Mackinnon M, White CP. The life expectancy of persons with cerebral palsy. Developmental Medicine and Child Neurology 1995; 37:567-576.

3. Evans PM, Evans SJW, Alberman E. Cerebral palsy: Why we must plan for survival. Archives of Disease in Childhood 1990; 65:1329-1333.

4. Strauss DJ, Shavelle RM, Anderson TW. Life expectancy of children with cerebral palsy. Pediatric Neurology 1998; 18:143-149.

5. Strauss DJ, Shavelle RM. Life expectancy of adults with cerebral palsy. Developmental Medicine and Child Neurology, in press.

6. Roberts, AH. Severe Accidental Head Injury. London: Macmillan, 1979.

7. Strauss DJ, Shavelle RM. Long-term survival of children and adolescents after traumatic brain injury. Archives of Physical Medicine and Rehabilitation, in press.

8. DeVivo MJ, Stover SL. Long-term survival and causes of death. In: SL Stover, JA DeLisa, GG Whiteneck (Eds.), Spinal Cord Injury, pp. 289-316. Gaithersburg MD: Aspen, 1995.


David Strauss, PhD, is a Fellow of the American Statistical Association and Professor of Statistics at the University of California, Riverside. He has frequently provided expert testimony on life expectancy in Britain, Canada, and the United States. He is director of the UC Riverside Life Expectancy Project, which specializes in the survival and life expectancy of persons with disabilities such as cerebral palsy and traumatic brain or spinal cord injuries.

Robert Shavelle, PhD, is a Visiting Professor at UC Riverside and a member of the Life Expectancy Project.