What this study adds
This study developed a clinical risk score to predict the risk of serious GI complications in elderly by using data from the entire Korean population, and demonstrated good performance through external validation.
This study identified new risk prediction factors not included in the current guidelines: male sex, very old age, and concomitant use of selective serotonin reuptake inhibitors.
Compared to the risk classification system of the current guidelines that count the number of risk factors, this newly developed risk score model included the concomitant use of a GPA as an offset factor, which enabled to capture expanded cases where use of proton pump inhibitors or H2-receptor antagonists was not sufficient to offset serious GI complications.
1. Introduction
Nonsteroidal anti-inflammatory drugs (NSAIDs) are one of the most frequently used medicines to treat musculoskeletal and rheumatic diseases [1]. NSAID use is a common cause of gastrointestinal (GI) bleeding, and mortality from GI bleeding has been reported to be 5–10% worldwide [2].
Several risk factors for GI injury in patients taking NSAIDs have been reported, including a history of GI ulcer, older age, and concomitant use of low-dose aspirin (ASA), other antiplatelet agents, anticoagulants, and corticosteroids [3]. Older age is a major risk factor for serious GI complications. In Italy, the prevalence of NSAID use among the elderly was estimated at 24.7% [4], and in the United States, 40% of the elderly people were prescribed with NSAIDs at least once a year [5]. The hospitalization rate for GI complications due to NSAIDs was 12 per 1,000 person-years in the elderly compared to < 1 per 1,000 person-years in the population aged < 50 years [5].
In 2014, a model for predicting the risk of upper GI bleeding in NSAID users was developed and verified [6], and a model for predicting the incidence rate of upper GI bleeding in NSAID users using a case-control study was developed [1]. There was a report on the risk score calculated for major toxicity, including adverse cardiovascular events, major GI events, acute kidney injury, and death in NSAID users from the PRECISION trial data in 2019 [7]; however, the outcome of this study was not specific for GI complications.
Although GI bleeding in the elderly is a burden due to the high mortality and disability rates [4], there is no risk stratification scheme to quantify the risk factors for GI complications in elderly NSAID users. Most geriatric focused criteria regarding the appropriate use of medication recommend the concomitant use of a gastroprotective agent (GPA) with chronic NSAID use [8]. However, GPA use might not eliminate the risk of serious GI complications in high-risk situations [8]. Few studies have included the use of a GPA in the prediction of GI complications.
Identifying high-risk patients in the elderly population would be helpful in preventing serious GI complications; therefore, developing a predictive risk score is important. We aimed to develop a risk prediction score to identify high-risk patients for severe GI complications in the elderly using NSAIDs and to validate it externally using nationwide claims datasets.