Our aim in this audit was to determine whether intensive rehabilitation post-operatively influenced length of stay and readmission rates for patients undergoing primary total knee arthroplasty.

In September 2007, a dedicated weekend physiotherapy service was set up in our Trust for patients following joint arthroplasty at a cost of £30,000 per annum. A prospective audit was conducted over two six-month periods, before and after the introduction of this service, including 202 and 240 patients respectively. Patient demographics including ASA grade and strict inclusion and exclusion criteria were used. The effect of anaesthetic type on post-operative pain control was also reviewed. Chi-squared and Mann-Whitney tests were used to analyse non-parametric data.

In the second cohort, with intensive rehabilitation, a statistically significantly higher number of patients were discharged within seven days of admission (64% vs 36%, p<0.01). This was despite there being a significantly higher number of patients with high ASA grades 3-4 in this cohort (37% vs 27%, p<0.05). The median length of stay in the second cohort was seven days compared to eight in the first cohort. There was a slight increase in rate of readmission within the second cohort but this was not statistically significant. We found that the addition of a femoral nerve block significantly reduced post-operative pain.

We concluded that an annual financial saving to the Trust of approximately £118,000 could be made by the addition of an additional dedicated physiotherapist in our unit. Patients can be safely discharged sooner with intensive rehabilitation and may benefit in the longer term by improved knee function.

In patients with an ankle fracture initial delay to operation because of time constraints is often prolonged because swelling precludes surgery for some days. We made use of a year long prospective audit of 2000 trauma patients to analyse the effect of delay to surgery on length of stay in ankle fracture patients.

One hundred and fifty patients were admitted with an ankle fracture. One hundred and twenty nine were operated on. The median (inter quartile range) time to surgery was 3 (2–5) days. Twenty six patients got to theatre within 24 hours. For those who didn’t get to theatre within 24 hours the median time to surgery was 4 days.

For the group as a whole there was a poor correlation between wait for surgery and length of stay (Pearson = 0.6). For the 98 patients under the age of 60 there was a significant relationship (Pearson co-efficient = 0.85). Fifty per cent of those under 60 were discharged within 48 hours of their surgery. The number of co-morbidities was different between the under and over 60s. The over 60s had a median (inter quartile range) of 2 (1–5) co-morbidities, compared to 0 (0–1) in those under 60.

Patients under 60 with an ankle fracture are generally medically fit. If 90% of such patients had their fractures fixed within 24 hours the median post operative length of stay for all ankle fractures in this population would fall from 7 to 3 days and the number of bed days saved would be 400 a year. The length of stay in patients over the age of 60 is more related to their associated co-morbidities than their time to surgery.

A prospective audit was carried out to analyse the relationship between time to surgery, number of co-morbidities and length of stay in 357 consecutive patients operated on for a fractured neck of femur.

One hundred and thirty five patients were operated on within 48 hours (group 1), 129 between 48 and 96 hours (group 2) and 93 patients after 96 hours (group 3). The mean (std dev) age was 77.2 (12.5) years in group 1, 79.8 (9.9) years in group 2 and 79.2 (9.4) in group 3. There were 93 (69%) females in group 1, 99 (77%) in group 2 and 67 (72%) in group 3. The number (%) admitted from home was 85 (63) in group 1, 81 (63) in group 2 and 73 (79) in group 3.

In the 30 patients with no co-morbidities there was a strong relationship between wait for surgery and length of stay. In these patients the median length of stay increased from 8.5 days in group 1 to 21 days when in group 3. In the 187 patients with one or two co-morbidities the relationship was present but weaker. The median length of stay increased from 16 days in group 1 to 21 days when in group 3. In the 140 patients with 3 or more co-morbidities there was no relationship between wait for surgery and length of stay. Median length of stay was 23 days in group 1 and 21 days in group 3.

This data from a large consecutive group of patients suggests that the fit patient with a hip fracture benefits from early surgery with a shorter length of stay. Those with multiple co-morbidities have their length of stay determined by their medical condition.