Back
Acute kidney injury (AKI) in people with chronic kidney disease (CKD) is an important driver of CKD progression and is associated with higher mortality and the need for kidney replacement therapy initiation. We investigated the incidence and characteristics of adult (>18 years old) clients accessing the public healthcare sector in the City of Cape Town with acute-on-chronic kidney disease (A-on-CKD) detected by a bespoke AKI algorithm from routinely collected administrative and serum creatinine (SCr) data collated by the Western Cape Provincial Health Data Centre (PHDC) in 2017–2021.
The algorithm detected changes in SCr consistent with Kidney Disease: Improving Global Outcomes-defined AKI. Episodes were limited to 7-day periods, to align with the delineation between AKI and acute kidney disease. The baseline SCr (the lowest value within 7 days or median 8–365 days) was fixed for the duration of the episode and the baseline period was reset after the index episode, for subsequent detected AKI. Time at-risk for A-on-CKD began on the date of CKD ascertainment if CKD was ascertained 2017 – 2021 or on 1 January 2017, when the AKI algorithm was activated, for CKD ascertained before 2017. Time at-risk ended on the date of in-facility death, if recorded, the date of last facility engagement or the date of the first AKI episode start, whichever came first. Recovery assessment was made at the most recent SCr test before the episode end: complete recovery (≤20% of the baseline SCr), partial (21–49%), nonrecovery (≥50%) and uncaptured (no test following the index SCr triggering AKI).
The cumulative incidence of at least one A-on-CKD episode was 11,798/63,085 (18.7%; mean age 66.6 years [SD 14.0], 61.8% female). Of those with CKD ascertained prior to 1Jan2017, the incidence rate was 74.8 per 1000 person-years at-risk (95% CI 73.1, 76.5) and of those with CKD ascertained on or after 1Jan2017: 86.3 (96% CI 83.7, 88.9). The KDIGO stage at presentation was KDIGO stage 1 for 61.4% of AKI, stage 2: 21.1% and stage 3: 17.6%. The first-ever episode was community-acquired (CA) in 78.9% and 21.1% were hospital-acquired (HA). Recovery after CA-AKI was complete in 13.4%, partial 9.6%, non-recovery 25.0% and uncaptured in 52.1% of individuals. For individuals with HA-AKI, there was greater complete (19.4%) and partial (15.1%) recovery. Otherwise, 32.7% had non-recovery by 7 days and 32.9% had no further SCr testing after the start of the HA-AKI episode. In-facility death occurred in 41.5% of individuals with uncaptured recovery of hospitalised CA-AKI and 47.2% of HA-AKI.
Describing the profile of clients with A-on-CKD in the City of Cape Town is novel since the PHDC is unique to the province of the Western Cape, and this information is important given the limited availability of nephrology services and KRT in the country. These results can be leveraged to advocate for the expansion of such resources locally. A-on-CKD was common, and despite the majority being stage 1, recovery was poor and in-facility mortality high. Significant internal migration, dual public-private healthcare structure in South Africa and assumptions about which individuals with CKD were still actively accessing public healthcare, and therefore at-risk of A-on-CKD, were limitations arising from the use of these routine health data.