105 Coding accuracy for severe preeclampsia requiring magnesium

Abstract

To determine ICD-10 accuracy in setting of MgSO4 seizure prophylaxis and examine whether clinical phenotype and SPE morbidity differs with correct vs. incorrect ICD-10 assignment. Retrospective chart review of ICD-10 codes in women with SPE requiring MgSO4 from 1-1-19 to 12-31-19. The study population was identified by MgSO4 release by the inpatient pharmacy. Inclusion criteria: age 18-50, >20 weeks, singleton or twins, treatment with MgS04. Exclusion criteria: gestational hypertension (GHTN), preeclampsia without severe features, MgSO4 for other indications, and missing data. Chi Square test and Fisher’s exact test were used for categorical variables. Wilcoxon Rank-Sum Test was used for continuous variables. 400 women received MgSO4. 22 women were excluded: 20 individuals received MgSO4 for other indications, 2 had missing coding data. Of the remaining 378: 272 (72.0%) had SPE and 106 (28.0%) had chronic hypertension (CHTN) with superimposed SPE. Of the 272 women with SPE, 243 (89.3%) had a correct code and 29 (10.7%) an incorrect code. 52 (49.0%) and 54 (51.0%) women with superimposed SPE had correct and incorrect codes, respectively. The most frequent incorrect code was CHTN with superimposed SPE (O11.4, 12.9%). Aside from diabetes – higher in incorrectly coded group – demographic, clinical characteristics, and severe maternal morbidity were not significantly different between groups. There were no maternal deaths. ICD-10 code assignment was more accurate for SPE than for superimposed SPE (Table). Although incorrect coding was not associated with increased disease severity or complications, ICD-10 coding was inaccurate in over 20% with SPE and nearly half with superimposed SPE. Preeclampsia research cannot rely on ICD-10 codes alone.