Letter

Standardised mortality ratios

Neither constant nor a fallacy

BMJ 2009; 338 doi: 10.1136/bmj.b1748 (Published 29 April 2009)
Cite this as: BMJ 2009;338:b1748

Access to the full text of this article requires a subscription or payment. Please log in or subscribe below.

  1. David I Ben-Tovim, director, clinical epidemiology unit1,
  2. Richard J Woodman, senior lecturer (biostatistics)2,
  3. Paul Hakendorf, senior epidemiologist1,
  4. James Harrison, director, AIHW national injury surveillance unit2
  1. 1Flinders Medical Centre, Bedford Park, SA 5042, Australia
  2. 2Faculty of Health Sciences, Flinders University, Bedford Park, SA 5042
  1. david.ben-tovim{at}health.sa.gov.au

    Mohammed and colleagues doubt the value of hospital standardised mortality ratios.1 They describe a possibility of increased bias consequent on a “constant risk fallacy” occurring if the effects of risk factors used in model adjustment process are assumed to be constant across hospitals in the subsequent hospital grouping, when they are not. Using data from four UK hospitals, Mohammed and colleagues showed …

    Access to the full text of this article requires a subscription or payment

    Subscribe

    Article access

    Article access for 1 day

    Purchase this article for £20 $30 €32*

    The PDF version can be downloaded as your personal record

    * Prices do not include VAT

    Rapid responses

    THIS WEEK'S POLL

    BMJ most popular