Article Text

Download PDFPDF

Hospital success rate: a multidimensional indicator of prognosis
  1. A Mason,
  2. M J Goldacre,
  3. H McGuinness
  1. Unit of Health-Care Epidemiology, Department of Public Health, University of Oxford, Oxford OX3 7LF, UK
  1. Correspondence to:
    Dr M J Goldacre, Unit of Health-Care Epidemiology, Department of Public Health, University of Oxford, Oxford OX3 7LF, UK;

Statistics from

There is worldwide interest in using routinely collected statistics to construct indicators to measure the outcome of hospital care. Typically, interest in any particular condition tends to focus on the review of comparative performance as assessed by individual measures such as mortality, emergency re-admission, or excessive lengths of stay. There is thus a set of indicators to review.

Consideration should also be given to the development of multidimensional outcome indicators, derivable from routinely collected data, that summarise the outcomes of a hospital admission and can be used to provide prognostic information. With the universal implementation of the NHS number in England, it will be possible to link data about hospital admissions for each individual with their subsequent admissions and with mortality data. National indicators based on linked data are now starting to be published.1 Using linked data, “success” can be considered in various ways and, as an illustration, in this study it has been defined at 90 days after admission as being alive, being out of hospital, and not having been re-admitted after discharge. We have calculated success rates for two conditions—stroke and fractured femur—highlighted as key priorities in the national strategic framework for older people.2

Routine data linked in the former Oxford health region were used (population 2.5 million) for the period 1 January 1994 to 31 March 1999. The denominators were emergency admissions for people aged 65 years and over with a principal diagnosis of stroke (ICD-9 431–434, and 436; ICD-10 I61–I64) or fractured neck of femur (ICD-9 820, 821.0 and 821.1; and ICD-10 S72.0, S72.1, S72.2 and S72.9). An admission was defined as a continuous stay in hospital, regardless of any change in consultant or hospital. Thus, admissions linked by a transfer were counted as a continuous stay. For each condition a second admission for the same condition within 90 days of the first was counted as a re-admission rather than a new fracture or stroke. The status of the patients was measured within 90 days of admission and classified as (a) dead, (b) re-admitted, (c) still in hospital with no break in inpatient care, or (d) none of these. If a patient had both a re-admission and died, only the death was included in the results. If a patient was readmitted more than once, he/she was counted once as a person re-admitted. The results presented in table 1 show that, at 90 days, the hospital success rate was 39 per 100 patients with stroke and 65 per 100 patients with fractured neck of femur.

We suggest that work might be undertaken in each clinical specialty to derive dimensions of outcomes from linked routine data. Key issues include the choice of conditions to be considered, the choice and definition of adverse events following care for the conditions (for example, whether to restrict analysis of re-admissions to those for certain types of complication); and time frames (for example, 30 or 90 days). If the indicator is to be used for making comparisons, consideration will have to be given to weighting the adverse events to reflect their relative importance. Whether weighted or not, we suggest that the components of the indicator are presented alongside the multidimensional measure.

Once the definition of “success” has been agreed for a given condition, it will be necessary to evaluate its usefulness in informing doctors and patients about the prognosis. However, any “success” indicator derived from routine statistics will be only a partial measure of outcome and will need to be supplemented by other clinician-assessed or patient-assessed measures.

Table 1

Outcome at 90 days following admission for stroke and fractured neck of femur: number and percentage of patients in each outcome category


View Abstract


  • Alastair Mason and Henry McGuinness are funded by the Department of Health as part of its funding of the National Centre for Health Outcomes Development. The views expressed in this letter are those of the authors and not necessarily those of the Department of Health. The Unit of Health-Care Epidemiology is funded by the South East Regional Office of the NHS Executive.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.