Is primary care a patient-safe setting? Prevalence, severity, nature, and causes of adverse events: numerous and mostly avoidable

BACKGROUND

Knowing the frequency and characteristics of adverse events is key to implementing actions that can prevent their occurrence. However, reporting systems are insufficient for this purpose and epidemiological studies are also required. Currently, the reviewing of clinical records is the gold standard method for knowing the frequency and characteristics of adverse events.Research on adverse events in primary care setting have been limited and primarily focuses on specific types of events (medication errors…) or patients. Large studies that search for any kind of adverse event in all patients are scarce.This study aimed to estimate the prevalence of adverse events in the primary care setting and their characteristics.

METHODS

Setting: all 262 primary healthcare centres in the Madrid region (Spain) during the last quarter of 2018.Design: cross-sectional descriptive study.Eligible population: subjects over 18 years of age who attended medical consultation over the last year (N=2,743,719). Randomised sample stratified by age.Main outcomes: age, sex, occurrence of an adverse event, number of consultations in the study period, avoidability, severity, place of occurrence, type of event, and contributory factors. The clinical records were reviewed by three teams, each composed of one doctor and one nurse trained and with expertise in patient safety.The SPSS software package (version 26) was used for the statistical analyses.

RESULTS

The evaluators reviewed 1,797 clinical records. The prevalence ofadverse events over the study period was 5.0%[confidence interval (CI) 95%:4.0%‒6.0%], with higher values in women (5.7%;CI95%:4.6%‒6.8%;P=0.10) and patients over 75 years of age (10.3%;CI95%:8.9%‒11.7%;P<0.001). The overall occurrence per hundred consultations was estimated to be 1.58% (CI95%:1.28%‒1.94%).Of the detected adverse events, 71.3%(CI95%:62.1%‒80.5%) were avoidable. Additionally, 60.6%(CI95%:50.7%‒70.5%) were categorised as mild, 31.9%(CI95%:22.4%‒41.4%) as moderate, and 7.4%(CI95%:2.1%‒12.7%) as severe. Primary care was the occurrence setting in 76.6%(CI95%:68.0%‒85.2%) of cases. The overall incidence of adverse events related to medication was 53.2%(CI95%:50.9%‒55.5%). The most frequent types of AE were prescription errors (28.7%;CI95%:19.5%‒37.9%), followed by drug administration errors by patients (17.0%;CI95%:9.4%‒24.6%), and clinical assessment errors (11.7%;CI95%:5.2%‒18.2%). The most common contributory factors were those related to the patient (80.6%;CI95%:71.1%‒90.1%) and tasks (59.7%;CI95%:48.0%‒71.4%).

CONCLUSION

A high prevalence of adverse events (1 in 66 consultations) was observed, which was slightly higher than those reported in similar studies. About three out of four such events were considered to be avoidable and one out of 13 were severe. Prescription errors, drug administration errors by patients, and clinical assessment errors were the most frequent types of adverse events.