A retrospective cohort study assessing the safety of post-hip fracture IV zoledronate given in renal impairment
Abstract
Introduction
Most hip fractures follow a fall. NOGG guidance recommends IV zoledronate first-line post hip fracture. Following a 2023 British Isles consensus statement, zoledronate is increasingly given to patients with renal impairment. This service evaluation assessed the safety of post-hip fracture IV zoledronate given in renal impairment.
Method
In a retrospective cohort study of sequential hip fracture admissions over 6-months, data were extracted from electronic medical records using a standardised template, regarding acute phase reaction (APR) and acute kidney injury (AKI). APR was defined as a temperature >37.5oC and/or heart rate >100 without infection; AKI as a serum creatinine increase ≥26 within 48 hours, or ≥50% within 7 days. Logistic regression assessed associations with AKI/APR.
Results
Of 317 adults [69.1% (n=219) female, median age 85 years (IQR 79-90)] 56(18.3%) had a CrCl <35. Overall, 197 (62.1%) received 5mg zoledronate, including 9 of 14 with CrCl 30-35. Median creatinine clearance (CrCl) at zoledronate administration was 55 (IQR 45-66) ml/min. 107/197 (54.3%) were on diuretics, and 31/107 (29.0%) had these held pre-zoledronate infusion.
16.8% (n=33/197) experienced an APR; normal renal function compared to impaired renal function increased the odds of APR (OR=3.19 [95%CI: 0.17, 59.6]); which was attenuated by age adjustment (OR=2.13 [0.17, 58.54). Where diuretics were paused pre-zoledronate, APR was still more common (OR 8.88[1.68, 46.9]) compared to those with diuretics maintained.
144 of 186 survivors (77.4%) had follow-up bloods available. Nine (6.25%) had an AKI, one of whom (11.1%) had had an initial CrCl 30-35, four (44.4%) had had CrCl 35-60 and four (44.4%) CrCl>60 (p=0.47).
Conclusion
Post-zoledronate AKI affected 6.3% of patients; but was no more common in those with CrCl 30-35, supporting the British Isles consensus statement. Post-zoledronate APR affected 17% of patients, slightly lower than some reports, potentially explained by slower administration with paracetamol co-prescription.