Impact of physician group practice in the operations, quality of care, and service satisfaction in the non-urgent section of an emergency department in a tertiary hospital in the Philippines: a mixed methods study | International Journal of Emergency Medicine

Design and population

The study utilized a sequential, explanatory, mixed-methods design; utilizing a review of GP and IC records, and a survey of patients. The quantitative component described the characteristics of the patient populations and ED operations, compared patient demographics and their assessment of the ED, and determined the financial components relative to GP and IC. Meanwhile, the qualitative component identified the subjective perception of the patients from both GP and IC, complementing the results from the quantitative component. The qualitative data were analyzed using the patient experience framework [14]. The study was conducted in the ED’s non-urgent section, which we defined as the Canadian Triage Acuity Scale categories 4 and 5 (patients with stable conditions) based on the patient’s chief complaints, modifiers, and physiologic parameters [15]. Both GP and IC concurrently handled non-urgent ED patients. For this study, the IC consists of consultants from varied specialties, including emergency and family medicine, providing patient care on 12-hour shifts two to three times a week and bound by individual annual contracts offering flexibility in terms of schedules. Most of the IC physicians have at least 10 years tenure and with multiple affiliations with other hospitals and ambulatory clinics, which may influence their breadth of experience and availability. In contrast, the GP is comprised specifically of occupational medicine physicians through a signed contract. The ED contracts GP to deliver 24/7 coverage for non-urgent cases, with the physicians working 12-hour shifts approximately two to three times per week. Most of the GP physicians are early in their careers, with fewer than five years of practice, classified as young professionals by age and service tenure.

Study setting

The study was conducted in the non-urgent section of the ED of a private tertiary hospital located in Metro Manila, Philippines with the highest ED patient census in the country. The hospital is for profit, with a capacity of 521 beds and 5,000 staff catering mostly to adult patients in middle- to upper-income classes [13]. Historically, the hospital ED engaged with ICs to render clinical services for the department. However, in 2021, at the height of the pandemic response, the sustainability of providing 24/7 service became a challenge, which prompted the ED to adopt a new business model by engaging with physician GPs. This arrangement ensures 24/7 clinical services in the non-urgent area of the ED. The current study banked on this arrangement as the ED is now being served by both GPs and ICs. The review of records included data from the non-urgent section of the ED from September 1, 2021 to December 31, 2022. Meanwhile, the patient survey was implemented on July 27 to October 29, 2023.

Study measures and procedures

For the historical data analysis, all available and complete data were gathered from the ED’s records (patient census, length of stay, ED physician’s attendance), electronic medical records (patient’s presenting condition, disposition, 72-hour unplanned return visits, mode of payment, medical orders, procedural orders), financial records (operational, medical and procedural costs), and customer service reports (complaints, ED satisfaction ratings, ED citations). Meanwhile, for the survey on non-urgent ED patients, a previously validated 31-item tool with questions on patient demographics, perception, ED accessibility, and descriptive satisfaction was used [9]. The survey was administered to an equal number of patients serviced by GP and IC. Specifically, the survey was offered by a dedicated research assistant to one in every four available non-urgent patients (based on the registration listing) during their consultation waiting period until the target sample size was achieved for each group. The minimum sample size for the survey (n = 310, 155 per group) was calculated at a power of 0.95 based on the effect size determined from a previous study [16] using G*Power software (HHU, Germany). All collected data were checked for completeness prior to analysis.

Analysis

A univariate descriptive analysis was utilized to describe the operational and patient characteristics of the two physician groupings (GP vs. IC). Meanwhile, all survey data was coded and interpreted based on a previous analysis workflow [9]. The Mann-Whitney U test was used to compare GP and IC physicians on the non-parametric comparison of the continuous measures of patient perception, ED accessibility, and satisfaction. An ANCOVA test compared the two physician groups on ED metrics, patient information, financial data, and customer service feedback. Month was included as a covariate alongside practice type to adjust for the temporal nature of the data. All levels of significance were set at alpha = 0.05, statistical encoding was done in a spreadsheet software, and analysis was conducted using jamovi (Version 2.5) and SPSS ver.22 (IBM SPSS, USA).

In addition, a qualitative analysis was conducted to the descriptive codes collected from the patient survey, following the patient experience framework [10, 13]. Specifically, individual codes were grouped based on commonality. The authors developed and agreed upon themes relating to ED service and physicians from the common code groups [17]. Disputes in themes were resolved through repeated discussions until a consensus was reached. A sequential explanatory analysis was used to integrate the data from the quantitative satisfaction survey and the qualitative themes [18].