NEW ZEALAND HEALTH INFORMATION SERVICE

Home > publications

This report is the result of comments and questions received following the release of the report Fracture of Neck of Femur Services in New Zealand 1998/99. There were two main objectives of this study:

• to use current administrative data (the mortality and morbidity data within the National Minimum Dataset) and statistical techniques to produce quality-of-care indicators for fracture-of-neck-of-femur services in New Zealand hospitals for clinicians.
• to evaluate the usefulness of current data collections for assisting clinicians with decision-making and with identifying areas for improvement.

The full methodology for this analysis is detailed in Appendix 1 of the report. The key statistical analysis performed was stepwise and logistic regression. Survival analysis tools developed by the Finnish Cancer Registry were used to analyse survival rates.

The study included 3131 patients aged 65 years and over with a hip fracture who were treated in public hospitals from 1 July 1999 to 30 June 2000. Out of that number, 2859 patients received surgical treatment and 272 were managed non-operatively.

Mortality

The findings of this study showed that 27 percent of patients died within 12 months of their injury. As the expected number of deaths estimated on the basis of patient’s age, sex and calendar time was roughly 10 percent, the actual burden of hip fractures was 17 percent. Mortality rates at one month and six months were 8 percent and 20 percent respectively. There was also a significant difference between sexes. Mortality for male patients was 37 percent, compared to 24 percent for females, although males were generally younger (median age 84 vs 87 years).

Patients treated surgically had a significantly lower mortality rate than those treated conservatively. The difference between the two treatment options was most obvious during the initial hospital episode (4 percent vs 20 percent).

Deaths occurred predominantly in hospital, either during the initial/procedure episode or during subsequent readmissions due to complications. Approximately one third of deaths occurred in the community.

Causes of mortality

The injury was not solely responsible for the high mortality rate. The combination of the injury with the patient’s age, sex, co-morbidity and case management were all significant factors.

The following predictors of death were identified as relevant by both stepwise and logistic regression modelling:

• age (being older)
• sex (being male)
• severity of co-morbid conditions (in particular, cardiac and repiratory)
• delay in time to surgery.

Delay in time to surgery, either because of pre-existing co-morbidity or because of non-medical reasons, was generally associated with increased mortality. The need to delay surgery may often be a reflection of a patient's general condition and measures instituted to reduce the anesthetic and operative risk. However, this study found that the medical condition was not exclusively responsible for delay. There is strong evidence that the number of patients operated on during the weekends is significantly lower than expected given the admission day. This was most likely to have been caused by the unavailability of surgery resources. It appears that patients who sustain a fractured hip are commonly treated with a low priority.

There was also considerable variation across the New Zealand hospitals in both the time from injury to admission and the time from admission to operation. Although the former time is not always recorded accurately, the findings of this study suggest that delay of surgery of more than two days after an injury is associated with increased mortality. This may indicate that the optimal time from injury to operation is approximately 48 hours and that all patients with a relatively stable medical condition should undergo surgery within that timeframe.

Clinical practice

Approximately 9 percent of patients in New Zealand did not proceed to surgery, either by patient choice or surgeon’s decision. Non-operative care was mostly associated with trochanteric fracture and co-morbid conditions. Besides showing a higher incidence of common cardiac and respiratory complications, patients cared for conservatively showed more nervous system and mental disorders than those treated surgically. The higher than expected rate of conservative treatment may reflect coding variations where avulsion fracture of the greater trochanter could not be separated from inter-trochanteric fracture.

There was variation in the surgical management of hip fractures across New Zealand hospitals. This is reflected in the variable proportion of internal fixation, hemiarthroplasty and total hip replacement. Some hospitals were identified as appearing to have preferences towards certain types of operation. However, it is always at the discretion of the surgeon to make the decision according to the individual circumstances.

Available as a zipped Word file (387 kB).

Published: 2002
48 pages, A4