Comprehensive review of ICD-9 code accuracies to measure multimorbidity in administrative data

Abstract

Background: Quantifying the burden of multimorbidity for healthcare research using administrative data has been constrained. Existing measures incompletely capture chronic conditions of relevance and are narrowly focused on risk-adjustment for mortality, healthcare cost or utilization. Moreover, the measures have not undergone a rigorous review for how accurately the components, specifically the International Classification of Diseases, Ninth Revision (ICD-9) codes, represent the chronic conditions that comprise the measures. We performed a comprehensive, structured literature review of research studies on the accuracy of ICD-9 codes validated using external sources across an inventory of 81 chronic conditions. The conditions as a weighted measure set have previously been demonstrated to impact not only mortality but also physical and mental health-related quality of life. Methods: For each of 81 conditions we performed a structured literature search with the goal to identify 1) studies that externally validate ICD-9 codes mapped to each chronic condition against an external source of data, and 2) the accuracy of ICD-9 codes reported in the identified validation studies. The primary measure of accuracy was the positive predictive value (PPV). We also reported negative predictive value (NPV), sensitivity, specificity, and kappa statistics when available. We searched PubMed and Google Scholar for studies published before June 2019. Results: We identified studies with validation statistics of ICD-9 codes for 51 (64%) of 81 conditions. Most of the studies (47/51 or 92%) used medical chart review as the external reference standard. Of the validated using medical chart review, the median (range) of mean PPVs was 85% (39-100%) and NPVs was 91% (41-100%). Most conditions had at least one validation study reporting PPV ≥70%. Conclusions: To help facilitate the use of patient-centered measures of multimorbidity in administrative data, this review provides the accuracy of ICD-9 codes for chronic conditions that impact a universally valued patientcentered outcome: health-related quality of life. These findings will assist health services studies that measure chronic disease burden and risk-adjust for comorbidity and multimorbidity using patient-centered outcomes in administrative data.

Definitions and Prevalence of Multimorbidity in Large Database Studies: A Scoping Review

“…Apart from inherent limitations of the coding systems used, ambiguity in patient record documentation and lack of clinical experience of coders affect coding accuracy [ 44 ]. While patient record review has been suggested as the best method to derive multimorbidity prevalence as it is not reliant on coding and data entry [ 15 , 16 ], large studies lack access to individual patient notes or direct accounts of patients’ conditions. Errors in coding subsequently implicate the accuracy of research using data from large databases [ 14 ].…”

“…This is confounded by the impracticability to go through individual patient notes or obtain a direct account of the patient’s conditions. It has been suggested that patient record review is the best way to collect information about multimorbidity prevalence, as it is not reliant on coding and data entry but rather gathers data from the entire patient record [ 15 , 16 ]. However, this is not feasible in large database studies.…”

“…Compared to smaller databases, standards of governance are generally required to ensure diagnoses are reliably coded in large datasets. Some challenges in using larger databases include the accuracy of the coding system used, and the reliability of coding as hospital records may be coded by non-physicians and may not reflect the actual codes from the physicians' perspectives [14,15]. This is confounded by the impracticability to go through individual patient notes or obtain a direct account of the patient's conditions.…”

Use of antipsychotic drugs and cholinesterase inhibitors and risk of falls and fractures: self-controlled case series

“…Many major disease diagnoses in the database have been validated by previous studies, including ischaemic stroke, 29 epilepsy, 30 hypertension, 31 diabetes, 31 hyperlipidaemia, 31 coronary artery disease, 31 atrial fibrillation, 31 heart failure, 32 Parkinson’s disease, 33 major neurocognitive disorders, 33 schizophrenia, 34 bipolar disorder, 34 and depression. 34 The diagnosis codes for osteoarthritis, 35 osteoporosis, 35 cataract, 35 falls, 36 and fractures 35 have not been validated but were selected based on previous studies and from the expert opinions of a psychiatrist and a geriatrician. Cholinesterase inhibitors and antipsychotic drugs are mostly reimbursed by the National Health Insurance programme in Taiwan, meaning that most prescription records have been captured.…”

Accuracy of administrative database estimates of national surgical volume: Solid organ transplantation in the National Inpatient Sample

“…1 The NIS has been employed to investigate assorted research questions in solid organ transplantation. [2][3][4][5][6][7][8][9][10][11][12][13][14][15][16] Although validation of diagnosis codes has been performed in numerous contexts, 17 less is known about surgical procedure coding. 18 National solid organ transplantation volume is closely tracked and validated, providing a unique opportunity to investigate accuracy.…”