RTI uses cookies to offer you the best experience online. By clicking “accept” on this website, you opt in and you agree to the use of cookies. If you would like to know more about how RTI uses cookies and how to manage them please view our Privacy Policy here. You can “opt out” or change your mind by visiting: http://optout.aboutads.info/. Click “accept” to agree.
Time between laboratory tests and acute liver and kidney injury diagnosis codes in CPRD
Are we doing it right?
Tormos, A., Rebordosa Garcia, C., Thomas, S. M., Plana, E., Rasouliyan, L., Pladevall-Vila, M., Farsani, S. F., Deruaz-Luyet, A., Brodovicz, K. G., & Perez-Gutthann, S. (2018). Time between laboratory tests and acute liver and kidney injury diagnosis codes in CPRD: Are we doing it right?Pharmacoepidemiology and Drug Safety, 27(S2), Article 988. https://doi.org/10.1002/pds.4629
Background: A 3‐month time window around time of diagnosis, commonly used in database studies to validate acute kidney injury (AKI) and acute liver injury (ALI), is based on clinical criteria that define chronic liver disease and chronic kidney disease (CKD), rather than on timing of acute events.
Objectives: To evaluate the validity of the 3‐month time window used as a first step to validate ALI and AKI in the Clinical Practice Research Datalink (CPRD).
Methods: Two cohorts (ALI and AKI) of adult new users of specific oral glucose‐lowering drugs (OGLD) for type 2 diabetes were identified in the CPRD, UK (August 2014‐October 2016). Patients with a previous diagnosis of ALI, AKI, or CKD were excluded, while those with previous abnormal laboratory values were not excluded. ALI and AKI cases were identified using diagnosis codes. Lab criteria were based on the Food and Drug Administration (FDA) definition for ALI or the adapted Risk, Injury, Failure, Loss, End‐stage (RIFLE) definition for AKI. We evaluated the period extending from 3 months before to 3 months after diagnosis. In addition, we evaluated the number of patients that fulfilled the lab criteria for ALI and AKI, irrespective of the presence of a diagnosis code.
Results: In the ALI cohort, among 14 805 new users, 64 potential cases were identified based on diagnosis codes; 6 of them fulfilled the lab criteria. Of the 6 patients, 5 fulfilled the lab criteria within 7 days after diagnosis and 1 within 31‐90 days. Seven patients fulfilled the lab criteria but did not have a diagnosis code for ALI. In the AKI cohort, among 12 346 new users, 23 potential cases were identified based on diagnosis codes; 7 patients fulfilled the lab criteria. Of the 7 patients, less than 5 fulfilled the lab criteria within 7 days after diagnosis, 1 within 16‐30 days. Two patients had a diagnosis code and fulfilled the lab criteria beyond the 3‐month criterion (more than 120 days). While not having a diagnosis code suggesting AKI, 151 patients fulfilled the lab criteria.
Conclusions: In the CPRD database, most cases (69%) fulfil the lab criteria within 1 week of the diagnosis code questioning the relevance of using a 3‐month time windows in such studies. A higher number of patients fulfilled the lab criteria suggesting AKI without a diagnosis code. These patients may have chronic liver or kidney disease where CKD was defined only by diagnosis codes.
