OBJECTIVES: SISPRO, the official Colombian Ministry of Health administrative database, which reports more than 500 million annual patient contacts, has been used to estimate prevalence of many diseases . We explored this database to establish different diagnoses that have been associated, during a five-year period, with ischemic stroke, and to quantify the magnitude of their statistical association.

METHODS: We recorded every patient 50 years of age or older that during the five-year period 2017-2021 had a diagnosis of ischemic stroke (ICD-10 codes i630-i669), as well as all subjects diagnosed with hypertension (i10x), diabetes (e10-e14), obesity (e660-e669), dyslipidemia (e780-e785), sleep apnea (g473) or atrial fibrillation (AF) (148x). For each of these latter conditions we estimate prevalence rate, using age- and gender-adjusted general population (all those that have contacted the health system over the five-year period) as well as the prevalence within ischemic stroke patients. With this information, we estimated prevalence ratios (PR) for each condition.

RESULTS: During the five-tear period, 12,848,169 persons age 50 or over (females 7,063,941, 55.0%) were attended by the Colombian healthcare system. Of these, 105,309 (females 53.167, 50.5%) had a diagnosis of ischemic stroke. Total patients diagnosed, ischemic stroke cases, and PR for each condition of interest, were, respectively: hypertension 5,112,234, 78,500, 3.06; diabetes 1,655,636, 25,254, 1.98; obesity 901,131, 6,751, 1.34; dyslipidemia 1,731,326, 11,509, 1.05; sleep apnea 216,508; 3,758, 2.72; and AF 98,201, 6,014; 6.53. PR was slightly higher for AF in females (6.74 vs 6.17), and higher for males than females with hypertension (3.52 vs 2.82).

CONCLUSIONS: Influence of risk factors for ischemic stroke has been normally established through cohort studies, like Framingham, or case-control studies, like INTERSTROKE. With this analysis from a large country-wide administrative database we use real world data as a form of triangulation, as well as providing some sort of face validity to our national database.