Little data exists on predicting the actual outcome of patients with fracture neck of femur when aged over 90. This group represents a complex of medical problems and where a delay in surgery can impact on patient recovery. In this study we evaluated the POSSUM scores at time of admission and time of surgery. We aim to define the actual mortality and morbidity of this group, if the possum had any predictive value, and any correlation with outcome.

132 patients over 90 with a fracture where followed from 2005–7, and a control cohort were followed up in 2005–6. A collection form was prepared to collect standard data on physiological status, with a standard scoring system on admission (Ortho possum), at the time of operation, comparing their progress and clinical outcome post-op. It also recorded co-morbidities and other outcomes. Statistical analysis was conducted using SPSS.

132 patient notes were reviewed and 130 patients in the control group. 5 had no surgery and the average age was 93 (90–103) vs. 76 in the control group. The majority of over 90’s were admitted from home by ambulance (n=99); and the cause of the fracture was recorded as a fall (n=68). 74 patients at admission were using a stick or a frame (24 were independent). Only 2 patients were on warfain.

At the point of admission the physiological POSSUM score on average was 23.48 (18–44) and at surgery it was 23.52 (16–38). This meant that the predicted mortality increased from 0.103 to 0.104. The average time to surgery was 1.5 days (0–12 days). However delays in surgery increased the POSSUM score and higher Possum scores were correlated with increased number of complications (p> 0.002), increased time to mobilisation (p> 0.003), and reduced mobility as compared to admission at day 15 and longer hospital admissions (p> 0.005).

In hospital mortality was 0.068 with a higher total POSSUM score prior to surgery of 36.29 for these patients compared to those patients who died after discharge. 35 patients died in total at 2 years post discharge (36%). Of these patients those within the 30 day mortality post discharge was 0.087 with a higher Total POSSUM score of 28.55 compared with the 120 day mortality post discharge of 0.194 with Total POSSUM score of 27.55; predicted mortality for the whole group was 0.28 using the Possum score (actual 0.27). Of the 35 patients that died 22 had higher Possum scores at surgery than admission.

In summary we found that there was no significant difference in the mortality and morbidity in the over 90’s fracture neck of femur group than the control. The Possum scoring system over predicted overall mortality and morbidity. Our results indicate a dedicated team to deal with these patients may well be of benefit to improve surgical Possum scores and outcome.