This algorithm builds off prior phenotyping work from Pacheco & Thompson available in the PheKB phenotype "Diverticulosis and Diverticulitis" as well as the manuscripts from Joo et al (2023)(1) and De Roo et al (2023) (2) . The objective is to approximate diverticular disease severity from the electronic medical record into groups of asymptomatic diverticulosis, mild diverticulitis, and severe diverticulitis.
This combines CPT codes, ICD-9 and ICD-10 codes, setting of care, and temporal relationships between codes. Cases of diverticular disease are separated from controls of no diverticular disease using an adaptation of the non-natural language processing algorithm described by Pacheco & Thompson. Severe diverticulitis is defined as a diverticulitis code on the same day as a colectomy or percutaneous abscess drainage, or at least two inpatient admissions for diverticulitis. Mild diverticulitis is defined as diverticulitis treated in the outpatient setting or with no more than 1 inpatient admission. Asymptomatic diverticulosis is defined through exclusive application of diverticulosis codes among cases.
Reference:
(1) Joo YY, Pacheco JA, Thompson WK, Rasmussen-Torvik LJ, Rasmussen LV, Lin FTJ, Andrade M, Borthwick KM, Bottinger E, Cagan A, Carrell DS, Denny JC, Ellis SB, Gottesman O, Linneman JG, Pathak J, Peissig PL, Shang N, Tromp G, Veerappan A, Smith ME, Chisholm RL, Gawron AJ, Hayes MG, Kho AN. Multi-ancestry genome- and phenome-wide association studies of diverticular disease in electronic health records with natural language processing enriched phenotyping algorithm. PLoS One. 2023 May 17;18(5):e0283553. doi: 10.1371/journal.pone.0283553. PMID: 37196047; PMCID: PMC10191288.
(2) De Roo AC, Chen Y, Du X, Handelman S, Byrnes M, Regenbogen SE, Speliotes EK, Maguire LH. Polygenic Risk Prediction in Diverticulitis. Ann Surg. 2023 Jun 1;277(6):e1262-e1268. doi: 10.1097/SLA.0000000000005623. Epub 2022 Jul 25. PMID: 35876359.
