Secondary analysis of outcomes after 11,085 hip fracture operations from the prospective UK Anaesthesia Sprint Audit of Practice (ASAP-2)

Anaesthesia 2016;71:506–514

Presented by: Dr M. Adamson

Background

Lack of high quality, prospective evidence to support particular anaesthetic practices for patients with hip fracture. RCTs are difficult to perform and observational studies are mostly retrospective case series with significant data omissions or concerns about data quality.

The National Hip Fracture Database (NHFD) collects observational data from >95% of all new cases of hip fracture in the UK (except Scotland), which can be audited against national standards.

The Anaesthesia Sprint Audit of Practice (ASAP) project, published 2014, was a national snapshot audit embedded within NHFD, collated anaesthetic and peri-operative variables for patients undergoing hip fracture surgery.  ASAP-1 found ‘striking’ inter-hospital variation in anaesthesia care, reflecting the uncertainties about what methods of anaesthesia might provide the best outcome for older, frailer patients with comorbidities requiring surgical hip fracture repair.

By revisiting the data collected for ASAP-1 and linking it to outcome data held by the NHFD, ASAP-2 aimed to determine whether there were any statistically significant associations between peri-operative patient factors, anaesthetic factors and outcomes.

Design & Setting

Data was taken from the ASAP-1 database and NHFD database. The original data was collected prospectively in the ASAP-1 database and then each patient was followed up for outcomes in the NHFD. The data collection period was May 1 2013 to July 31 2013.

Subjects

Patients >60yrs old who had hip fracture surgery between May 1 and July 31 2013 at one of 182 UK hospitals that routinely contribute to NHFD.

Intervention

There was no specific intervention. All patients underwent hip fracture surgery and the following data was collected.

Patient variables on hospital admission: age; sex; comorbidities; ASA physical status; place of residence (home, sheltered, rehabilitation, residential home, nursing home, inpatient, other); independence (self-caring/home-help/sheltered/residential/nursing); and cognition (abbreviated mental test score).

Dates and times of hospital admission and surgery; the type of anaesthesia administered (general, nerve block, spinal); the type and quantity of intrathecal injectate; the seniority of operative surgical and anaesthetic personnel; intra- operative blood pressures; and signs of bone cement implantation syndrome.

Systolic and diastolic blood pressures recorded at two times: immediately before commencement of anaesthesia; and the lowest recorded intra-operative value. Mean arterial blood pressure calculated as the diastolic pressure plus one-third of the difference between systolic and diastolic pressures. 

Outcomes

Outcome measures were recorded:

  • Postoperative cognition and residential destination on discharge.
  • Mortality at 30 postoperative days
  • Mortality 5 days after hospital admission.
  • Post-operative length of stay (the time from surgery to discharge from the acute hospital) 

Results

The NHFD recorded 16,904 operations between 1st May and 31st July 2013.

  • 11,130 (67.5%) were audited by ASAP-1
  • 45 duplicate records
  • 11,085 analysed in ASAP-1 and 2

The proportion of cases for which the anaesthetic technique was categorised identically by the National Hip Fracture Database and the ASAP data collectors ranged from 35% for general anaesthesia to 88% for general anaesthesia supplemented by epidural, peripheral nerve or surgical site injection of local anaesthetic.

No differences were found in outcomes (Mortality 5 day/30 day, length of stay, deterioration in cognition, independence or residential status) when comparing type of anaesthetic given.

Patients cared for by a combination of consultant or specialist surgeon and anaesthetist were on average one year younger (than patients cared for by other grades) but more often were ASA physical status 4 or 5 (p < 0.0001 for both). There were no differences in outcomes apart from survivors spent 0.7 days (17 h) less in hospital after surgery by a combination of consultant or specialist surgeon and anaesthetist, compared with other grades.

Mortality at five and 30 postoperative days was associated with lower intra-operative blood pressures.

  • The OR (95% CI) for 30-day mortality was 0.992 (0.986–0.998) for each 5 mmHg increase in systolic pressure, p = 0.0075, and 0.985 (0.977–0.992) for each mmHg increase in mean pressure, p < 0.0001. Similar relationships were seen for five-day mortality odds ratios. These relationships persisted when adjusted for Nottingham Hip Fracture Score.
  • An intra-operative systolic blood pressure below 85 mmHg compared with higher systolic pressures was associated with higher mortalities: five-day mortality 64/3,062 (2.1%) vs. 78/ 7,427 (1.1%) respectively, p = 0.017; 30-day mortality 181/3,062 (5.9%) vs. 338/7,427 (4.6%) respectively, p = 0.013.

The relative fall in systolic blood pressure was weakly correlated with more sub- arachnoid bupivacaine: r2 0.10 and 0.16 for hyperbaric and isobaric bupivacaine, respectively. A 20% relative fall in systolic blood pressure correlated with 1.4 ml hyperbaric bupivacaine 0.5% and 1.5 ml isobaric bupivacaine 0.5%.

Conclusions

No association between mortality and anaesthetic technique

No association between mortality and

  • Day of the week
  • Time of day
  • Grade of surgical or anaesthetic personnel

Mortality at 5 days and 30 days postoperatively were associated with lower intra-operative blood pressures

Lower intra-operative blood pressures were weakly associated with higher volumes of intrathecal local anaesthetic.

Strengths

  • Prospective data
  • Large numbers of patients
  • Data point completion rates (>90% of most fields)
  • ASAP data input by anaesthetists about anaesthesia likely to be more accurate than NHFD

Weaknesses

  • Observational study therefore can only say associations not causation
  • Results only as good as data input – incomplete data sets therefore affect results
  • Only included 67.5% patients who fractured their hip during the 3 month study period
  • The 30 day mortality rate for this period was lower than the annual average (5.1% vs 8.0%) which may be a reflection of some missing dat from hospitals that didn’t participate
  • Some of the hospitals eligible that weren’t included had higher than  average mortality rates and hence useful data may have been missed

Implications

  • Mode of anaesthesia makes no difference……its more about how you conduct anaesthesia
  • Avoid hypotension during anaesthesia for hip fracture surgery

Potential for impact

  • Only associations so not major impact but large numbers and suggestions are seem pragmatic