门诊老年患者潜在不适当用药对全因住院时长和住院药品费用的影响

The impact of exposure of different numbers of potentially inappropriate medications among old outpatients on all-cause hospitalization expenditures and length of stay

目的:研究门诊老年人使用潜在不适当用药(PIM)对患者全因住院时长和住院费用的影响。方法:回顾性分析北京市医保数据库2016年7-9月数据。纳入的患者年龄≥65岁,应用Beers Criteria 2015版识别PIM。暴露PIM的定义为在3个月内被处方至少一个PIM。结局事件定义为患者在暴露于PIM或非PIM 14 d后(≥14 d)发生住院。通过门诊数据中患者ID匹配住院数据的患者ID,抓取发生住院的患者信息,计算这些患者的住院时长和住院药品费用。结果:研究共纳入506 214位患者,中位年龄74岁(范围65~105岁),50.10%为女性。门诊老年人服用PIM后发生住院的患者共5 200人(1.03%)。服用PIM=0,PIM=1,PIM=2、PIM≥3后住院天数分别10.00 d(7.00,14.00),11.00 d(7.00,15.00),11.00 d(8.00,14.00)和11.00 d(7.00,14.00),组间差异有统计学意义(P=0.01);服用PIM=0,PIM=1,PIM=2,PIM≥3后住院费用分别13 488.42元(8 926.00,20 914.36),14 325.67元(9 484.81,22 626.31),14 201.07元(9 274.79,21 673.86)和14 214.64元(9 716.09,22 098.23),组间差异无统计学意义(P=0.20)。经多因素分析后,PIM暴露数量与住院时长和住院费用无相关性(P>0.05);年龄≥75岁、胆碱能评分(ARS)≥2、糖尿病、就诊次数5~8次可增加住院时长(P<0.05);查尔森合并症指数(CCI)≥3分、二级医院就诊、就诊次数5~8次、药品使用种类数5~9种增加住院药品费用(P<0.05)。结论经多因素校正后,本研究发现门诊老年患者暴露于不同数量PIM不增加住院时长和住院药品费用,为评估PIM对住院医疗资源的使用提供参考。

OBJECTIVE To study the effect of potential inappropriate drug use(PIM) on the length of all-cause hospitalization and hospitalization expenses of elderly outpatients.METHOD Retrospective cohort study used data from Beijing’s Longitudinal Health Insurance database.The study included a community-dwelling sample of adults aged 65 and older from July to September 2016.PIMs were identified based on the 2015 Beers Criteria^(?).PIM exposure was defined as taking at least one PIM within study period.Clinical outcome were defined as hospitalization after 14 days(≥14 d) of exposure to PIM.Patient ID in the outpatient data file matched with the patient ID in the inpatient data files to determine all-cause hospitalizations.Medication expenditures and length of stay of hospitalization were calculated.RESULTS 506 214 patients were included,with a median age of 74 years(range 65-105 years),and 50.10% of them were women.The numbers of patients exposed to PIM who ended with hospitalization were 5 200(1.03%).The length of stays of hospitalization were 10.00 days(95%CI:7.00,14.00)(PIM=0),11.00 days(95%CI:7.00,15.00)(PIM=1),11.00 days(95%CI:8.00,14.00)(PIM=2) and 11.00(95%CI:7.00,14.00)(PIM≥3),respectively.The difference between groups was statistically significant(P=0.01).After taking PIM=0,PIM=1,PIM=2 and PIM≥3,the medication expenditure of hospitalization were ￥13 488.42(95%CI:8 926.00,20 914.36),￥14 325.67(95%CI:9 484.81,22 626.31),￥14 201.07(95%CI:9 274.79,21 673.86) and ￥14 214.64(95%CI:9 716.09,22 098.23),respectively.There was no significant difference between the groups(P=0.20).After multivariate analysis,there was no correlation between PIM exposure and length of stay and medication expenditures of hospitalization(P>0.05).Increased length of stay were associated with age≥75 years,anticholinergic risk scale(ARS score)≥2,diabetes and number of visits(P<0.05).Increased medication expenditures of hospitalization were associated with Charlson comorbidity index(CCI)≥3,secondary hospitals,5-8 outpatients visits,and 5-9 medication use.CONCLUSION With the increase of PIM exposure,the length of stay of hospitalization were increased.The result provided a reference basis for policy makers and health care providers to evaluate the impact of PIM on utilization of inpatient medical resources.