RESUMO
Introduction Certain patient groups perceive specific barriers to accessing primary care, resulting in increased emergency department (ED) use for non-emergency conditions. There is evidence coming from other countries that homeless people are treated differently in accessing emergency services. Examination of ED wait time by demographic characteristics provides pertinent information to identify subgroups that are more subject to the consequences or causes of access block and delayed treatment. Methods We analyzed five years of Emergency Department Information System (EDIS) visit records of the largest tertiary care center in Atlantic Canada. The wait time from triage to seeing a physician was the outcome, housing status was the main exposure, and age and gender were the predictors. Quantile regressions were carried out to identify the influence of homeless visits in meeting the Canadian Association of Emergency Physicians (CAEP) wait time benchmarks for each Canadian Triage and Acuity Scale (CTAS) level. The classification and regression tree method was used to quantify and classify the demographic subgroups contributing to wait time disparities across CTAS levels. Results Homeless visit median wait times that exceeded the three-hour CAEP benchmark were significantly longer for urgent (by 40 minutes, CI: 25-55), semi-urgent (by 31 minutes, CI: 17-45), and non-urgent (by 57 minutes, CI: 25-89) than acuity level-matched domiciled visit wait times. At the 50th percentile, one-hour benchmark homeless triaged as semi-urgent waited (median=20 minutes, CI: 12-28) longer, and no other triage-level differences were found at this benchmark. Homeless emergent-level visits that exceeded the three-hour benchmark were 28 minutes, on average, shorter than domiciled patients of the same acuity level. Homeless females above 40 stayed the longest for non-urgent care (mean=173 minutes), 82 minutes longer (p=0.0001) than age-gender-acuity level-matched domiciled patients. Homelessness was the most prominent ED wait time classifier for non-urgent, ED visits. Overall, homeless patients triaged as CTAS-5 waited 30 minutes longer (p=0.0001) than domiciled patients triaged as CTAS-5. Homeless male 16-20-year-olds waited the shortest time of 72 minutes. Conclusion Homelessness-related wait time disparities exist in the low acuity non-urgent-level ED visits more than in the other levels, supporting the theory that lack of primary care access is a driver of ED use in this group. Our acuity level analysis supports that homeless people of a certain age (older) and gender groups (female) wait longer than their age-gender-matched domiciled patients to be seen by a physician in low acuity level presentations. Given the pattern of the homeless being seen earlier or statistically similar in emergent-level visits compared to matched domiciled patients and that 16-20-year-old homeless males were seen on average within 72 minutes (the shortest mean wait time reported for the triage level CTAS-5), we decline the notion of discrimination at the study site ED. If homeless patients' non-urgent needs were met elsewhere, pressure on the ED to meet benchmarks might be reduced.
RESUMO
Introduction Canadian emergency departments (EDs) are struggling under the weight of increased use by a growing population of elderly patients; those who lack proper housing; and those who lack family physicians to provide primary care. The Canadian Foundation for Healthcare Improvement projected a possible ED service utilization increase in Canada at a rate of 40% over three decades. This calls for local-level information on the time trends to understand demographic and temporal variations in the different geographical locations in the country. This study sought to identify and quantify acuity level-based per capita ED visit annual time trends for the 10-year period of 2006-2015 (by age, gender, and housing status). The aim is to provide detailed information on the time trends for demographically targeted ED planning locally. The lengthy record of data allows examining the changing directions in different time segments. Material and methods Administrative data from the largest emergency department in Halifax (Nova Scotia, Canada) was analyzed. Per capita adult ED visit rates (EDVR) based on Canadian Triage Acuity Scale (CTAS), age, gender, and housing status were analyzed. Trends in the age-gender-based standardized rates using 2011 census city population data were also estimated in order to account for the population increase in the city. Results No study in Canada has documented the possibility of flattening the escalating ED visit trend by maintaining an annual declining trend in low-acuity-level visits or documented a threshold rate of decline to be maintained. This study observed that the annual linear per capita non-homeless EDVR increment trend (328/year, CI:245-411, per 100,000) for all-acuity-level visits - noted for a ten-year period - would become stable when low-acuity-level CTAS4-5 visit declining trends (427/year, CI:350-503 and 121/year, CI:79-163, per 100,000) - noted for the period of 2012-2015 - were maintained at the same magnitude and direction. Alarming annual emergent (high acuity level of CTAS2) EDVR increase equivalent to 335/year (CI:280-391, per 100,000) was noted for all combined visits, from all age, gender, and housing groups visits. The highest incremental rate noted among above-50-year-olds (521/year, per 100,000, 95% CI:433-608) was neither driven by overall increasing population census numbers nor by increasing aging population numbers. We found statistically similar age-gender standardized rates (294/year, CI: 207-382) for all ED visits and (316/year, CI:261-372) for CTAS2 level visits, when adjusted for annual population increase. Homeless ED visits did not contribute to the overall ED visit incremental trend. The highest annual homeless increment rate was shown for <30-year-old group high acuity CTAS-2 level visits (219/year, CI:193-246, per 100,000). Conclusion Neither the city population increase nor increased homeless visits contributed to ED visit annual per capita incremental trends in the city of Halifax. The increasing trend was chiefly driven by high-acuity-level visits by >50-year-old patients. Our findings suggest one way to make this escalating ED visit rates stable in the future is by maintaining the declining semi-urgent and non-urgent visit trends at the same rates estimated within the years 2012-2015. These findings highlight the potential directions for ED services planning to keep up with the growing demand for high-acuity-level ED services by the aging population.
