Another Study of the Effect of Increasing Specialist Availability on Hospital Mortality Rates

News Blog readers will know that a study associated with CLAHRC WM is evaluating the effect on excess mortality among patients admitted over the weekend. This prospective study, led by CLAHRC WM Leadership Fellow Julian Bion, is making use of a ‘natural experiment’ as hospitals increase consultant presence at different rates. A similar study has examined the effect of gradual increase in intensivist staffing in American intensive care units.[1] An association at baseline between intensivist staffing and mortality was found. But hospitals “that adopted intensivist staffing during the study period did not substantially improve their mortality rates.” The relative risk was 0.96 (0.90 – 1.02) on the difference-in-difference measure. An observer with a symmetrical prior centred on a 4% difference in the difference would, of course, be able to cling onto that 4% effect size. If all adverse events were reduced in proportion, would this be cost-effective at the nominal US willingness-to-pay threshold for a year of healthy life of about $100,000. The CLAHRC WM Director feels a modelling study coming on. But here is a back-of-the-envelope calculation. ITU mortality in the US is about 12%.[2] This could be reduced by 4% to about 0.5%, or about five lives in a thousand ITU admissions. That amount to around 50 life years if the mean life expectancy of a life saved is 10 years. Fifty life years represents an (undiscounted) expected benefit of $5,000,000. Any less and the extra intensivists would represent poor value for money.

— Richard Lilford, CLAHRC WM Director


  1. Nagendran M, Dimick JB, Gonzalez AA, Birkmeyer JD, Ghaferi AA. Mortality Among Older Adults Before Versus After Hospital Transition to Intensivist Staffing. Med Care. 2016; 54(1): 67-73.
  2. Zimmerman JE, Kramer AA, Knaus WA. Changes in Hospital Mortality for United States Intensive Care Unit Admissions from 1988 to 2012. Crit Care. 2013; 17: R81.

One thought on “Another Study of the Effect of Increasing Specialist Availability on Hospital Mortality Rates”

  1. Richard – another fascinating blog but I think you need a bigger envelope… Assuming the transition to intensivist staffing does not change physician to patient ratios, we need to know about the salaries and on-costs of the physicians replaced by intensivists during the transitions as well as those of the intensivists themselves (according to the 2015 Medscape survey, Critical Care physicians are around the centre of the salary distribution across specialties). It’s the incremental cost of intensivists (compared with who they replace) that matters under the above assumption; and it is this cost that needs to be less than $5m (before discounting) per 1,000 admissions.

    A US Health Resources & Services Administration report to Congress on the Critical Care workforce reports a mean length of stay in the ICU of around 4 days per patient, so that’s around 4,000 patient days for our 1,000 admissions. The report cites the COMPACCS study which suggests each patient requires 45 minutes of intensivist care per day; equivalent to around 3,068 hours or, assuming 6 hours of direct patient care delivered per day by each intensivist, 510 days. Moving to intensivist staffing therefore needs to cost less than $9,800 per day – which would actually buy 10 additional critical care physicians. Someone clinging to their 4% relative reduction in mortality would therefore suggest the move would represent good value for money (again, before discounting).

    And if, like me, you found applying a difference in difference result to reduce 12% “by 4% to about 0.5%” (sic; it should have been TO about 11.5% or OF about 0.5%) a little confusing, then please support efforts to help us all understand the difference between absolute and relative risks and percentage changes!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s