Tuesday, 26 June 2012

Life Expectancy after Delay in Diagnosing Cancer - Part Two

In Part One I set out how the decision in Gregg-v-Scott directed litigators away from using chances of five or ten year survival as a means of assessing the impact of a negligent delay in diagnosing cancer, and towards the use of median life expectancy.

The median of a group is not the same as its mean. The mean life expectancy of a group of women with breast cancer is the average of their total expected life spans. The median life expectancy (MLE) is the age which each is likely to reach but beyond which they are not likely to live. The way I think of it is that the 50th person in the line of 100, ordered according to life expectancy, is the median.

In a case of negligent delay in diagnosing cancer we need the MLE following actual diagnosis and the MLE had earlier diagnosis been made. Each will depend on the features of the cancer at the relevant time, the treatment which would have been offered and which the claimant would have elected to undergo and, perhaps, other factors such as age. Let's disregard the impact of co-morbidities for the purposes of this blog, but factors such as smoking history, diabetes or hypertension might well be relevant. Once you have MLE for each date then you can calculate the reduction in life expectancy due to the delay (not forgetting to take into account the years that have passed during the delay).

I have been extolling the benefits of MLE for some time now. It was therefore pleasing to hear Professor Gordon Wishart, Breast Surgeon, speak at the annual AvMA conference in Brighton this month (June 2012). He said that for the purposes of litigation, life expectancy should be looked at in term of numbers of years rather than chances of survival. He has helped develop the PREDICT model which can be used to predict life expectancy following breast cancer. I have conducted many breast cancer cases where there has been a dispute about whether the Nottingham Prognostic Index or Adjuvant-on-Line gave the more accurate prognosis. PREDICT is a recent development which may become the favoured prognostic model amongst experts - time will tell. It has the advantage over AoL in being based on data of British women treated here, and certainly takes account of all modern treatment options.  It can also be used to give MLE.

With MLE the Court can quantify future claims, knowing (on the balance of probabilities) over what period such future claims should be made. Claims such as earnings in the lost years can be assessed and, if the claim is under the  Fatal Accidents Act, the Court will have evidence of how long the deceased would have been likely to live without the negligent delay in diagnosis.

However, whilst there is a huge amount of data in relation to survival following breast cancer, there maybe insufficient data to allow for calculations of MLE in relation to less common cancers. Further, it is not clear whether the data is sufficiently robust to allow calculation of MLE beyond say 15 or 20 years. In several cases which I have conducted, expert evidence from a medical statistician has been utilised to assist in calculating MLE. I would be interested to learn of others' experience in these often difficult cases.

One general difficulty with the use of statistical evidence of life expectancy is that with every day that passes, the claimant has lived a day longer. Good news for the claimant of course, but sometimes patients "buck" the statistics. You may even have a claimant who, by the date of trial, has already lived longer than expected. What is his or her life expectancy then? On the other hand there may be claimants who die earlier than expected. Is it correct to assume that with earlier diagnosis they would also have died earlier than their MLE would suggest? Or is their premature death following actual diagnosis irrelevant to the prediction of their likely age at death following timely diagnosis? To what extent does a finding of metastatic spread after actual diagnosis affect the calculation of MLE following timely diagnosis? No easy answers but I hope to have sparked a few thoughts and ideas. Please share them.

See my profile on the Kings Chambers website. My twitter link is there also.



No comments:

Post a Comment