Surgical site infections following orthopaedic surgery: Statewide surveillance using linked administrative databases

In many institutions prospective surveillance programmes to monitor the incidence of post-operative infection can be diffi cult to implement due to limited human and technical resources. In addition, prolonged patient follow up, up to one year, may be required for implant surgery. Traditional methods of surveillance can be enhanced by using administrative databases to assist in case fi nding and facilitate overall surveillance activities. The aim of this study was to identify the incidence of surgical site infection (SSI) in patients who had undergone total hip replacement (THR) or total knee replacement (TKR) surgery in all Western Australian (WA) hospitals in 1999 using the Western Australian Health Services Research Linked Database. The database was used to identify patients who underwent THR or TKR surgery in 1999 using ICD–9-CM and ICD–10-CM codes. Of these, patients who had been given an infection diagnosis code plus the external cause ‘surgical operation with implant of artifi cial internal device’ codes (E878.1 in ICD–9-CM, Y83.1 in ICD–10-CM) were identifi ed. This allowed all patients to be followed for at least one year after surgery. Patients who died from other causes during the follow-up period were identifi ed from linked mortality data. A total of 1476 THR and 1875 TKR procedures was identifi ed from 21 WA hospitals (11 public, 10 private) during 1999. There were 169 infections identifi ed, giving an overall cumulative incidence of 5%, (95% CI 4.3–5.7) [THR (4.86%, 95% CI 3.77–5.95) and TKR (5.15%, 95% CI 4.15–6.15)]. Only 23 infections were coded specifi cally as ‘infection and infl ammatory reaction due to hip or knee replacement’. The remainder had been given alternative infection codes, but had also been given the external cause code ‘surgical operation with implant of artifi cial internal device’. There was a total follow-up time of 5012 person-years, giving an incidence rate of 33.72 infections per 1000 person-years. Most infections (96%) occurred within 1 year of surgery. Patients aged over 80 years experienced a signifi cantly higher rate of infection after THR, when compared to patients aged 80 or less (z-test, z=2.56, p=0.0150), but not for TKR (z=0.35, p=0.726). Eighty-fi ve patients (50.3%) developed a SSI during the same hospital admission as the surgical procedure, the remainder were reSurgical site infections following orthopaedic surgery: Statewide surveillance using linked administrative databases