Impact of primary care system interventions on healthcare worker satisfaction and intention to stay in the Philippines: a follow-up study

Abstract

Background

The Philippines has a shortage and uneven distribution of healthcare workers (HCWs). Job satisfaction is an important element to HCW retention and attracting new HCWs into the health system.

Objective

This study measured HCWs’ intent to stay and HCWs’ satisfaction after implementation of multiple interventions intended to strengthen the primary care system, and determine factors significantly associated with HCWs’ intent to stay.

Methodology

This is a serial cross-sectional study in urban, rural and remote primary care sites in the Philippines. All physicians, nurses, midwives, dentists, community health workers and support staff were invited to participate. Baseline HCWs’ intent to stay and satisfaction were obtained using a self-administered questionnaire prior to implementation of interventions. The same survey was again conducted in the years 2021 and 2022, corresponding to 5 and 6 years after initial implementation for the urban site, and 2 and 3 years for the rural and remote sites. We used multiple logistic regression to determine factors associated with intent to stay.

Results

There were 430 survey respondents (89.4% response rate) for year 2021, and 417 survey respondents (97.4% response rate) for year 2022. The urban and rural sites had significant increase in several HCW satisfaction domains, while the remote site had significant decrease in several HCW satisfaction domains. There was no significant difference in the intent to stay in the three sites. Factors that decreased intent to stay included length of employment, job involvement and employment as a nurse, while factors that increased intent to stay included job satisfaction, enjoyment and working in the urban site.

Conclusion

HCW satisfaction improved in the urban site and rural site, while HCW satisfaction declined in the remote site. Intention to stay of primary care HCWs did not significantly change.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• The Philippine healthcare system faces a myriad of challenges to the delivery of equitable and accessible healthcare for all Filipinos, particularly the shortage and uneven distribution of healthcare workers (HCWs).

• HCW job satisfaction is an important element to HCW retention and attracting new HCWs into the health system.

WHAT THIS STUDY ADDS

• This is a follow-up study describing HCW satisfaction in urban, rural and remote primary care sites in the years 2021 and 2022, up to 6 years after implementation of the interventions intended to strengthen primary care systems.

• HCW satisfaction improved in the urban site and rural site, while HCW satisfaction declined in the remote site.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Strong organisational support is needed when instituting healthcare system changes to promote HCW satisfaction, especially in resource-constrained settings.

Introduction

The Philippine healthcare system continues to face a myriad of challenges to the delivery of equitable and accessible healthcare for all Filipinos. One important challenge is the shortage and uneven distribution of healthcare workers (HCWs). The Philippines currently has 19.7 HCWs (referring to physicians, nurses, midwives and dentists) per 10 000 population, which are less than half of the WHO-recommended minimum density of 44.5 HCWs per 10 000 population.^{1 2}

Approximately 62 000 new health workers are produced annually in the Philippines, with a population of approximately 110 million.³ However, almost one-fifth of Filipino HCWs work in other countries. Among those who work in the Philippines, majority are concentrated in urban areas.⁴ Of the 17 regions in the Philippines, only 2 regions (National Capital Region and the Cordillera Administrative Region) meet the recommended number of doctors.¹ Furthermore, there is also an internal exodus of HCWs to jobs with better pay and working conditions, with only approximately 22% of HCWs in active service in the health sector.⁴

HCW job satisfaction is an important element to HCW retention and attracting new HCWs into the health system. HCW satisfaction is affected by several factors, including competitive pay, good working conditions, reasonable workload, opportunities for career growth, job autonomy and positive relationships with colleagues.⁵ High HCW retention and satisfaction translate to better patient care.⁶ Evaluating factors affecting HCWs’ intention to stay is important in strengthening human health resources and ensuring quality healthcare.

The Philippine Primary Care Studies (PPCS) Programme was launched in 2016 to develop and test strategies to strengthen primary care systems through multisectoral interventions. The interventions involved health financing through a primary care benefit package, community engagement and expansion of healthcare provider networks through partnerships with private laboratories and pharmacies. Human resource development was also supported through provision of HCW training, augmentation of health human resources, provision of performance-based financial incentives and implementation of a health information system. The PPCS pilot tested these interventions in three primary care sites: an urban, rural and remote site.

HCWs’ intention to stay and satisfaction before and 1 year after implementation of interventions were previously reported. Briefly, it was observed that there was significant improvement in perceived fairness in compensation among HCWs in the urban and rural sites. Intent to stay remained the same for HCWs in the urban and rural sites, but there was a decline in intent to stay in the remote site.⁷

Interventions implemented by the PPCS received full funding (amounting to 400 Philippine pesos (PhP) per capita per year) up to year 1 (year 2016–2017 for urban site, year 2019–2020 for rural and remote sites). Thereafter, only residual funds were available (25% of the total budget), and the timing of release of funds was very irregular. Funding was already depleted by year 2020 in the urban site and year 2022 in the rural and remote sites (see online supplemental appendix for detailed funding status). Moreover, the COVID-19 pandemic in 2020 drastically changed the health landscape, affecting both the provision of healthcare services and the health-seeking behaviour of patients.⁸ With these changes, it is necessary to re-evaluate the effect on HCWs’ intent to stay. This study aimed to (1) compare HCW satisfaction in the three sites before implementation of the interventions and in the years 2021 and 2022; (2) compare HCW intention to stay in the three sites during these said years; and (3) determine factors that are significantly associated with HCW intention to stay.

Methodology

Study design

This is a serial cross-sectional study that assessed HCW intention to stay and satisfaction in three pilot sites in the Philippines. Baseline HCW intention to stay and satisfaction were obtained using a self-administered questionnaire in September 2016 for the urban site and in January 2017 for the rural site and November 2017 for the remote site, prior to implementation of interventions.⁷

Interventions to strengthen primary care were implemented in the year 2016 for the urban site and 2019 for the rural and remote sites. These interventions included (a) health financing through provision of a primary care benefit package (400 per capita per year) that covered outpatient consultation services, laboratory and diagnostic procedures and medicine of up to Php2000 per patient per year (transportation allowance was also provided for patients in the remote areas); (b) health information system through development of an electronic health record (EHR) system; (c) HCW training; (d) HCW support through hiring of additional physicians and provision of performance-based financial incentives; (e) instituting healthcare provider networks with private laboratories and pharmacies; and (f) community engagement through information dissemination. The detailed description of these interventions is in the online supplemental appendix. More information on the PPCS Programme may be found in other published papers.^9–11

HCW intention to stay 1 year after implementation was previously described.⁷ This is a follow-up study assessing HCW intention to stay at years 2021 and 2022 compared with baseline. The survey was conducted in May–June 2021 and May 2022, which corresponded to 2 and 3 years after initial implementation for the rural and remote sites, and 5 and 6 years after initial implementation for the urban site.

Setting

This study involved three primary care sites in the Philippines. The urban site was at the University of the Philippines (UP) Health Service in Diliman which provided healthcare services to approximately 15 051 employees of UP Diliman. The rural site was in the Municipality of Samal in the Province of Bataan. This was a fourth-class municipality catering to 35 298 residents. The remote site was in the Municipality of Bulusan in the Province of Sorsogon. This site is considered a geographically isolated and disadvantaged area (GIDA), catering to 22 884 residents.

Study participants

All HCWs, including physicians, nurses, midwives and dentists, in the three primary care sites were invited to participate in the study. We also invited community health workers and support staff in the three sites to participate.

Variables and data measurement

This study used the same prevalidated and translated self-administered Stayers questionnaire used to evaluate HCW intention to stay at baseline and 1 year post-implementation.⁷ The questionnaire consisted of 18 items, with 5 items related to sociodemographic characteristics of the participants, 11 items related to HCW satisfaction and 2 items related to intent to stay.

For the 11 items related to HCW satisfaction, participants were asked to choose their level of agreement to each item using a 5-point Likert scale. The 11 items were classified as motivator factors and hygiene factors using Herzberg’s two-factor theory, also known as the motivator–hygiene theory. Motivator factors refer to elements that motivate employees to stay in their jobs, while hygiene factors refer to elements that satisfy basic needs of employees.¹² In the questionnaire, the five motivator factors included overall job satisfaction, workplace morale, recommendability, enjoyment and job involvement. The six hygiene factors in the questionnaire were manageable workload, access to supplies, equipment and medicines, job security and fairness of compensation.⁷

For the two items related to intent to stay, these items were multiple choice questions with five response options. The first item was a general question on intent to stay, with the five response options ranging from leaving the job as soon as possible, leaving the job within a year, leaving the job 1–2 years from now, leaving the job 3–5 years from now and staying in the job indefinitely. The second item applied to respondents who answered that they will leave the job soon and inquired where they plan to transfer to. Response options included staying with the same organisation but transferring to a different location, transferring to a different organisation, transferring to a job outside the health sector and leaving the country. The full questionnaire can be found in online supplemental appendix 1.

Sampling and survey distribution

Total enumeration was done for this study. The questionnaire was distributed through a virtual platform to all HCWs, including physicians, nurses, midwives, community health workers and support staff, in the three primary care sites. In case of non-response, the participants were followed up by the research team and encouraged to complete the short questionnaire.

Statistical analysis

Study data were processed using Microsoft Excel. Sociodemographic characteristics were presented using frequencies and percentages for categorical variables, and means and SDs for continuous variables for respondents per site in the years 2021 and 2022.

Likert responses were scored as 1=strongly disagree, 2=disagree, 3=neutral, 4=agree and 5=strongly agree. Responses to the 5-point Likert scale were presented using median and IQR for baseline, 1 year after implementation, and years 2021 and 2022.

For the items related to intention to stay, responses were described using frequencies and percentages for each site. Responses were also dichotomised as HCWs intending to leave regardless of time period (includes leaving the job as soon as possible, within a year, 1–2 years and 3–5 years) and HCWs intending to stay indefinitely. Results were presented for baseline, 1 year after implementation, and years 2021 and 2022.

The 99.2% CI of the true median score of each factor was the basis of comparison among the three time periods: at baseline, and in the years 2021–2022. The 99.2% CI of a single proportion was used to compare intent to stay (as a dichotomous variable) across the three time periods. The 99.2% CI was used to adjust for multiple comparisons. Multiple logistic regression was used to determine factors that are significantly associated with intent to stay. ORs were adjusted for possible confounders, including sex, job description, length of employment, job experience and primary care site. Regression analysis was done separately for years 2021 and 2022. Data analysis was done using STATA V.14.0

Results

Study participants

We invited all 481 HCWs to participate in the study for the year 2021 and 427 for the year 2022. There were 430 survey respondents (89.4% response rate) for the year 2021, with 93 from the urban site, 151 from the rural site and 186 from the remote site. There were 417 survey respondents (97.4% response rate) for the year 2022, with 86 from the urban site, 153 from the rural site and 178 from the remote site. The flow diagram of the study participants is in the online supplemental appendix.

In the rural and remote sites, a great majority of the respondents were female. The most common job description of the respondents was community health workers. In the urban site, there were slightly more females than males. The most common job descriptions include administrative aides, nurses and physicians. Most of the respondents had previous work experience. The median length of employment in the primary care facility was 3.5–4.0 years for the rural site, 6.0–6.3 for the remote site and 13.9–15.0 years for the urban site (table 1).

Table 1.

Sociodemographic profile of study participants, n (%)

	Year 2021 N=430			Year 2022 N=417
	Urban site N=93	Rural site N=151	Remote site N=186	Urban site N=86	Rural site N=153	Remote N=178
Sex
Male	45 (48.4)	8 (5.3)	4 (2.2)	38 (44.2)	12 (7.8)	9 (5.1)
Female	48 (51.6)	143 (94.7)	182 (97.8)	48 (55.8)	141 (92.2)	169 (94.9)
Job description
Physician	11 (11.8)	1 (0.7)	3 (1.6)	13 (15.1)	4 (2.6)	1 (0.6)
Nurse	19 (20.4)	17 (11.3)	11 (5.9)	16 (18.6)	17 (11.1)	10 (5.6)
Midwife/nurse aide*	6 (6.5)	21 (13.9)	12 (6.5)	5 (5.8)	20 (13.1)	13 (7.3)
Community health worker	0	85 (56.3)	142 (76.3)	0	89 (58.2)	137(77.0)
Pharmacist	3 (3.2)	0	1 (0.5)	4 (4.7)	0	0
Medical technologist/radiation technologist	10 (10.8)	2 (1.3)	4 (2.2)	9 (10.5)	2 (1.3)	0
Administrative aide	21 (22.6)	0	10 (5.4)	19 (22.1)	7 (4.6)	4 (2.2)
Dentist	4 (4.3)	9 (6.0)	2 (1.1)	5 (5.8)	1 (0.7)	0
Technical aide	5 (5.4)	4 (2.6)	0	3 (3.5)	0	6 (3.4)
Others†	14 (15.1)	12 (7.9)	1 (0.5)	12 (14.0)	6 (3.9)	7 (4.0)
Job experience
First job	20 (21.5)	60 (39.7)	81 (43.5)	16 (18.6)	57 (37.3)	67 (37.6)
With previous job	73 (78.5)	91 (60.3)	105 (56.5)	70 (81.4)	96 (62.7)	111 (62.4)
Length of employment in years, median (IQR)	13.9 (22.0)	3.5 (9.2)	6.0 (8.5)	15.0 (21.9)	4.0 (8.0)	6.3 (8.1)

*Midwives for rural and remote sites, nurse aides for urban site.

†Includes ambulance drivers, sanitation inspectors, utility personnel, information technology personnel and food attendants.

HCW satisfaction through time in the three sites

Table 2 shows the scores to the domains related to HCW satisfaction at baseline, year 2021 and year 2022. Compared with baseline, there was a significant increase in workplace morale, access to equipment and perceived compensation fairness at year 2021. In the year 2022, access to equipment and perceived compensation fairness decreased to baseline levels, There was also significant decrease in enjoyment and access to medicine compared with baseline and year 2021.

Table 2.

HCW satisfaction scores from baseline to year 2022—median, range (99.2% CI)

Domain	Baseline N=356	Year 2021 N=430	Year 2022 N=417
Motivation factors
Job satisfaction	5, 2–5 (4.8 to 5.0)	5, 1–5 (4.8 to 5.0)	5, 2–5 (4.8 to 5.0)
Workplace morale	4.5, 1–5 (4.4 to 4.6)	5, 1–5 (4.8 to 5.0)*	5, 2–5 (4.8 to 5.0)*
Recommendability	5, 2–5 (4.9 to 5.0)	5, 1–5 (4.9 to 5.0)	5, 1–5 (4.9 to 5.0)
Enjoyment	5, 2–5 (4.8 to 5.0)	5, 1–5 (4.8– to 5.0)	4, 1–5 (3.9 to 4.1)†
Job involvement	5, 2–5 (4.9 to 5.0)	5, 1–5 (4.9 to 5.0)	5, 1–5 (4.9 to 5.0)
Hygiene factors
Manageable workload	4, 2–5 (3.9 to 4.1)	4, 1–5 (3.9 to 4.1)	4, 1–5 (3.8 to 4.2)
Access to supplies	4, 1–5 (3.8 to 4.2)	4, 1–5 (3.8 to 4.2)	4, 1–5 (3.8 to 4.2)
Access to equipment	3, 1–5 (2.8 to 3.2)	4, 1–5 (3.8 to 4.2)*	3, 1–5 (2.8 to 3.2)
Access to medicine	4, 1–5 (3.8 to 4.2)	4, 1–5 (3.8 to 4.2)	3, 1–5 (2.8 to 3.2)†
Perceived job security	4, 1–5 (3.8 to 4.2)	4, 1–5 (3.8 to 4.2)	4, 1–5 (3.8 to 4.2)
Perceived compensation fairness	3, 1–5 (2.8 to 3.2)	4, 1–5 (3.8 to 4.2)*	3, 1–5 (2.8 to 3.2)

Values in bold show statistically significant difference from baseline

*Significant increase compared to baseline

†Significant decrease compared to baseline

HCW, healthcare worker.

Table 3 shows the satisfaction scores categorised based on site of employment of the respondents. The urban site had a significant increase in job satisfaction, recommendability, job involvement, manageability of workload, and perceived job security at years 2021 and 2022 compared with baseline. Enjoyment at year 2021 significantly improved compared with baseline, but decreased to baseline levels at year 2022. Workplace morale significantly increased at year 2022 compared with baseline and year 2021.

Table 3.

HCW satisfaction scores per site from baseline to year 2022—median, range (99.2% CI)

Domain	Urban site			Rural site			Remote site
	Baseline N=36	Year 2021 N=93	Year 2022 N=86	Baseline N=141	Year 2021 N=151	Year 2022 N=153	Baseline N=179	Year 2021 N=186	Year 2022 N=178
Motivation factors
Job satisfaction	4, 3–5 (3.6 to 4.4)	5, 2–5 (4.7 to 5.0)*	5, 3–5 (4.7 to 5.0)*	5, 3–5 (4.7 to 5.0)	5, 1–5 (4.7 to 5.0)	5, 2–5 (4.8 to 5.0)	5, 2–5 (4.7 to 5.0)	4, 3–5 (3.8 to 4.2)†	4, 2–5 (3.8 to 4.2)†
Workplace morale	4, 1–5 (3.5 to 4.5)	4, 3–5 (3.7 to 4.3)	5, 3–5 (4.7 to 5.0)*	5, 3–5 (4.8 to 5.0)	5, 1–5 (4.7 to 5.0)	5, 2–5 (4.8 to 5.0)	5, 2–5 (4.7 to 5.0)	4, 1–5 (3.8 to 4.2)†	4, 2–5 (3.8 to 4.2)†
Recommendability	4, 3–5 (3.6 to 4.4)	5, 3–5 (4.7 to 5.0)*	5, 1–5 (4.7 to 5.0)*	5, 3–5 (4.8 to 5.0)	5, 1–5 (4.8 to 5.0)	5, 1–5 (4.8 to 5.0)	5, 2–5 (4.8 to 5.0)	5, 3–5 (4.8 to 5.0)	4, 1–5 (3.8 to 4.2)†
Enjoyment	4, 3–5 (3.6 to 4.4)	5, 1–5 (4.6 to 5.0)*	4, 1–5 (3.7 to 4.3)	5, 3–5 (4.8 to 5.0)	5, 1–5 (4.7 to 5.0)	5, 2–5 (4.8 to 5.0)	5, 2–5 (4.8 to 5.0)	4, 1–5 (3.8 to 4.2)†	4, 2–5 (3.8 to 4.2)†
Job involvement	4, 3–5 (3.6 to 4.4)	5, 3–5 (4.8 to 5.0)*	5, 1–5 (4.7 to 5.0)*	5, 3–5 (4.8 to 5.0)	5, 1–5 (4.7 to 5.0)	5, 2–5 (4.8 to 5.0)	5, 2–5 (4.8 to 5.0)	5, 1–5 (4.8 to 5.0)	4, 1–5 (3.8 to 4.2)†
Hygiene factors
Manageable workload	4, 2–5 (3.6 to 4.4)	5, 1–5 (4.6 to 5.0)*	5, 1–5 (4.7 to 5.0)*	4, 2–5 (3.8 to 4.2)	5, 1–5 (4.7 to 5.0)*	5, 2–5 (4.8 to 5.0)*	4, 2–5 (3.8 to 4.2)	4, 2–5 (3.8 to 4.2)	3, 1–5 (2.7 to 3.3)†
Access to supplies	4, 3–5 (3.6 to 4.4)	4, 2–5 (3.7 to 4.3)	4, 1–5 (3.7 to 4.3)	4, 1–5 (3.8 to 4.2)	5, 1–5 (4.7 to 5.0)*	5, 1–5 (4.7 to 5.0)*	4, 2–5 (3.8 to 4.2)	3, 1–5 (2.7 to 3.3)†	3, 1–5 (2.8 to 3.2)†
Access to equipment	4, 2–5 (3.4 to 4.6)	4, 1–5 (3.7 to 4.3)	4, 1–5 (3.7 to 4.3)	3, 1–5 (2.6 to 3.4)	4, 1–5 (3.7 to 4.3)*	4, 1–5 (3.7 to 4.3)*	3, 2–5 (2.7 to 3.3)	3, 1–5 (2.7 to 3.3)	3, 1–5 (2.7 to 3.2)
Access to medicine	4, 2–5 (3.5 to 4.5)	4, 1–5 (3.7 to 4.3)	4, 1–5 (3.6 to 4.4)	3, 1–5 (2.7 to 3.3)	4, 1–5 (3.7 to 4.3)*	4, 1–5 (3.7 to 4.3)*	4, 2–5 (3.8 to 4.2)	3, 1–5 (2.7 to 3.3)†	3, 1–5 (2.8 to 3.2)†
Perceived job security	4, 3–5 (3.6 to 4.4)	5, 2–5 (4.7 to 5.0)*	5, 3–5 (4.8 to 5.0)*	4, 1–5 (3.7 to 4.3)	5, 1–5 (4.7 to 5.0)*	5, 2–5 (4.7 to 5.0)*	4, 2–5 (3.8 to 4.2)	4, 1–5 (3.7 to 4.3)	3, 1–5 (2.8 to 3.2)†
Perceived compensation fairness	3, 1–5 (2.3 to 3.7)	4, 2–5 (3.7 to 4.3)	4, 1–5 (3.7 to 4.3)	3, 1–5 (2.7 to 3.3)	4, 1–5 (3.7 to 4.3)*	4, 1–5 (3.7 to 4.3)*	3, 2–5 (2.7 to 3.3)	3, 1–5 (2.7 to 3.3)	3, 1–5 (2.8 to 3.2)

Values in bold show statistically significant difference from baseline

*Significant increase compared to baseline

†Significant decrease compared to baseline

HCW, healthcare worker.

The rural site had significant increase in satisfaction relating to work hygiene factors, including manageability of workload, access to supplies, access to equipment, access to medicine, perceived job security, and perceived compensation fairness at year 2021 and year 2022 compared with baseline.

In contrast, the remote site had significant decrease in three out of the five motivation actors (job satisfaction, workplace morale, enjoyment) and two out of the five hygiene factors (access to supplies and access to medicine) at year 2021 and year 2022 compared with baseline. Recommendability, job involvement, manageability of workload and perceived job security significantly decreased in year 2022 compared with baseline and year 2021.

HCWs’ intent to stay through time in the three sites

Intent to stay from baseline to year 2022 is shown in figure 1. There was no significant difference in the intent to stay across the three time periods for each site and overall.

Figure 1.

Comparison of percentage of healthcare workers who intend to stay, from baseline to year 2022, in urban, rural and remote sites.

Factors affecting intent to stay

Factors affecting intent to stay are shown in table 4. Only length of employment was significantly associated with intent to stay for both year 2021 (adjusted OR 0.96, 95% CI 0.928, 0.988) and year 2022 (adjusted OR 0.96, 95% CI 0.929, 0.989). Intent to stay with the primary care facility is 4% lower for every year increase in length of employment. Simply put, the longer an HCW has been working in a primary care facility, the lower is the HCW’s intent to stay.

Table 4.

Regression model of factors affecting intent to stay

Variables	Year 2021		Year 2022
	Crude OR (95% CI)	Adjusted OR (95% CI)	Crude OR (95% CI)	Adjusted OR (95% CI)
Job satisfaction	1.32 (0.999, 1.748)	1.10 (0.653, 1.858)	1.72 (1.20, 2.47)	4.58 (2.302, 9.121)
Workplace morale	1.29 (0.978, 1.704)	0.91 (0.530, 1.548)	1.31 (0.880, 1.947)	1.11 (0.543, 2.252)
Recommendability	1.26 (0.917, 1.740)	0.59 (0.326, 1.073)	1.08 (0.751, 1.566)	0.92 (0.489, 1.737)
Enjoyment	1.58 (1.207, 2.058)	1.87 (1.199, 2.925)	1.44 (0.994, 2.087)	1.71 (0.814, 3.577)
Job involvement	1.17 (0.879, 1.565)	0.94 (0.580, 1.532)	0.75 (0.472, 1.194)	0.42 (0.198, 0.909)
Workload	1.340 (1.011, 1.776)	1.18 (0.770, 1.796)	0.62 (0.461, 0.840)	0.61 (0.363, 1.109)
Access to supplies	1.15 (0.945, 1.423)	1.04 (0.750, 1.455)	0.85 (0.655, 1.111)	0.85 (0.500, 1.433)
Access to equipment	1.16 (0.968, 1.393)	1.07 (0.806, 1.420)	0.92 (0.735, 1.159)	1.20 (0.803, 1.801)
Access to medicine	1.18 (0.96, 1.46)	1.07 (0.777, 1.475)	0.81 (0.628, 1.039)	0.81 (0.527, 1.257)
Perceived job security	1.38 (1.111, 1.720)	1.28 (0.951, 1.723)	0.61 (0.453, 0.826)	0.69 (0.433, 1.087)
Compensation fairness	1.05 (0.863, 1.281)	0.97 (0.721, 1.308)	0.76 (0.597, 0.968)	1.02 (0.702, 1.495)
Sex	1.48 (0.787, 2.769)	1.97 (0.804, 4.839)	2.44 (1.289, 4.605)	2.20 (0.940, 5.143)
Length of employment (in years)	0.99 (0.867, 1.009)	0.96 (0.928, 0.988)	0.97 (0.946, 0.987)	0.96 (0.929, 0.989)
Job description
Physician	0.50 (0.163, 1.538)	0.41 (0.111, 1.488)	0.68 (0.197, 2.384)	0.70 (0.159, 3.104)
Nurse	0.60 (0.377, 1.300)	0.46 (0.186, 1.147)	0.52 (0.204, 1.311)	0.19 (0.060, 0.614)
Midwife	1.16 (0.464, 2.910)	1.58 (0.508, 4.926)	4.65 (0.588, 36.846)	2.33 (0.250, 21.755)
Barangay health workers	1.35 (0.747, 2.427)	2.33 (0.972, 5.566)	1.95 (1.077, 3.539)	1.62 (0.693, 3.792)
Others*	Reference	Reference	Reference	Reference
Primary care site
Remote	Reference	Reference	Reference	Reference
Urban	1.30 (0.696, 2.444)	6.35 (2.097, 19.247)	0.394 (0.198, 0.786)	3.07 (0.914, 10.280)
Rural	0.97 (0.580, 1.611)	1.26 (0.610, 2.620)	0.510 (0.274, 0.954)	0.79 (0.312, 1.981)
Job experience	0.83 (0.513, 1.338)	0.85 (0.489, 1.477)	.0.73 (0.412, 1.297)	0.76 (0.384, 1.515)

In bold are the statistically significant scores based on the 99.2% CI.

*Includes medical and allied health professionals (pharmacists, medical technologists, radiology technicians) and administrative staff.

In the year 2021 regression model, enjoyment and working in the urban site were also significantly associated with intent to stay. Intent to stay is 1.87 times higher for each unit increase in enjoyment score (adjusted OR 1.87, 95% CI 1.199, 2.925). Intent to stay is 6.35 times higher for those working in the urban site compared with the remote site (adjusted OR 6.35, 95% CI 2.097, 19.247), although significant association was found only after adjustment was done in the analysis. In the univariate analysis, job involvement and site of employment showed no significant association with intent to stay.

In the year 2022 regression model, job satisfaction, job involvement and working as a nurse were also significantly associated with intent to stay. Intent to stay is 4.58 times higher for each unit increase in job satisfaction score (adjusted OR 4.58, 95% CI 2.302, 9.121). However, intent to stay is 58% lower for each unit increase in job involvement score (adjusted OR 0.42, 95% CI 0.198, 0.909). Intent to stay was also 81% lower among nurses compared with other positions, such as administrative staff, technicians, dentists and pharmacists (adjusted OR 0.19, 95% CI 0.060, 0.614).

Discussion

This study demonstrated no significant change in intent to stay in all three primary care sites in the Philippines despite implementation of interventions that intended to strengthen the primary care system from baseline (year 2016 for the urban site, year 2019 for the rural and remote sites) to years 2021 and 2022. This ranged from 65% to 90% throughout the study in all three sites.

Analysis of HCW satisfaction per site shows how inequity was inadvertently aggravated in the remote site after interventions were implemented, especially during the COVID-19 pandemic. There was a significant decrease in almost all domains of HCW satisfaction in the remote site, including job satisfaction, workplace morale, recommendability, enjoyment, access to supplies and access to medicine. In contrast, the rural site had significant improvement in perceived workload, access to supplies, access to equipment, access to medicine, job security and compensation fairness.

HCW satisfaction

The study results are consistent with our initial study comparing HCW satisfaction before and 1 year after implementation, which reported a decline in several factors in the remote site.⁷ This follow-up study showed that the decline in HCW satisfaction persisted until year 2022, 3 years after implementation. In contrast, the urban and rural sites had significant improvement in several factors. In most of these factors, the improvement was noted only by years 2021 and 2022.

Several studies have demonstrated that implementation of any organisational change is difficult, even if the change is intended to improve the organisation and its employees.^13–16 The biggest challenge is the attitude and beliefs of the employees towards the change, since employees often have negative and fearful perceptions of any proposed changes to their work environment. This phenomenon has been labelled as resistance to change (RTC), and is considered a normal response to perceived threat to the status quo.^{13 14} Perceived organisational support, which refers to how well an organisation is able to support its employees’ welfare, social needs and career development, can lower RTC.¹³ Moreover, being prepared for the instituted changes has been shown to be an important determinant of the positive perception of HCWs towards the change.¹⁵

In our study, the remote site, as a GIDA, faced the largest challenge in terms of availability and accessibility of human resources, laboratories, equipment, medicines and infrastructural support even before the implementation of interventions. These barriers may explain why the implemented interventions failed to lead to an improvement in HCW satisfaction. The resources in the remote site, as well as organisational support, were not large enough to mitigate RTC. Furthermore, the remote site experienced several challenges in the implementation of interventions, including the limited satellite providers available for this site. The 24 barangay health stations could not be connected under a single network. As a result, the full potential of the EHR system could not be realised, and patients in remote barangays needed to travel long distances to avail of services in their rural health unit. The remote site also faced shortage in HCW staff. The decline in HCW satisfaction persisted until 3 years post-implementation, which may be due to (1) persistence in RTC, (2) depletion of available funding for the instituted changes and (3) impact of the COVID-19 pandemic on HCW welfare.

In the urban and rural sites, the available resources to support the organisation prior and during the implementation of interventions were much greater compared with the remote site. The urban site had a good complement of HCWs, consisting of physicians, nurses and midwives. They also had an in-house network that facilitated the integration of the EHR. In the rural site, various healthcare programmes were already existent and well instituted prior to the implemented changes. The local government unit also supported the PPCS interventions by installing additional towers with long-distance routers to unify all 14 barangays under a single network. However, it is interesting that most of the factors related to HCW satisfaction only improved by year 2021.

According to the model by Kübler-Ross, the initial perception of health personnel to any institutional change is negative. Feelings of shock, anger, denial and fear often surface due to the disruption. As time progresses, the impact of the changes becomes positive, leading to exploration and rebuilding.¹⁶ This model may explain why significant improvement in various factors related to HCW satisfaction was more evident in the year 2021. In the initial evaluation of HCW satisfaction 1 year after implementation, it is possible that the HCWs were still in the disruption phase. Hence, there was no significant rise in most domains of HCW satisfaction.⁷ However, there was also no significant decline in HCW satisfaction due to mitigating factors that could have lessened the negative impact of changes, including strong organisational support and readiness for change. Despite the decrease in available funds in years 2021 and 2022, the positive impact on HCW satisfaction of the changes was observed. During these years, the positive impact of non-financial interventions such as the integrated EHR system, HCW training and community engagement may have persisted, leading to sustained improvement in HCW satisfaction.

HCWs’ intent to stay

In this study, factors that significantly affect intent to stay include length of employment, enjoyment, job satisfaction, job involvement, working in the urban site and working as a nurse. Job involvement, longer employment and working as a nurse were associated with lower intent to stay, while enjoyment, job satisfaction and working in the urban site were associated with higher intent to stay.

Several international studies have been conducted on determinants for job retention. A systematic review on factors that impact retention among HCWs reported that job satisfaction, career development and work–life balance were the major determinants for job retention among HCWs, consistent with the results of our study. Job satisfaction was described as the complex interaction between job demands and job resources.¹⁷

A study in Italy reported that HCWs working in provinces with perceived shortage of medical personnel had lower job satisfaction and intent to stay, compared with those in high-quality facilities.¹⁸ This is consistent with our study results that those working in the urban site predicted higher odds of intent to stay. The increase in job resources, including human health resources, in the urban site in our study may have been a positive factor contributing to higher intent to stay.

A study on nurses’ intent to stay in Turkey reported that the number of years working in an organisation predicted intention to stay. Specifically, the longer the years of employment, the higher was the intention to stay. The study authors explained that this was due to increased autonomy and control of nurses over their work practices with longer employment, leading to higher intent to stay.¹⁹ In our study, we observed the opposite effect—those with longer employment had lower intent to stay. A possible reason for this is that HCWs who have been working longer may possibly experience burnout and emotional exhaustion in our study, leading to lower intent to stay. More research on the existing mental and emotional health status of HCWs in primary care settings in the Philippines is needed to explore this hypothesis.

A review of well-being of different HCWs reported that physicians and nurses have higher perceived workload-related irritability compared with other HCWs. Moreover, nurses have higher adverse experiences (such as depression) in different types of disasters.²⁰ These may explain our study findings of lower intent to stay among nurses compared with other HCWs.

Limitations

One limitation of this study is the possible confounding effect of extraneous variables. The onset of the COVID-19 pandemic in the Philippines in March 2020 greatly altered the healthcare system demands and processes. Several studies have documented the negative impact of the pandemic on HCW satisfaction.^{6 21} Another limitation is the depletion of available funds to sustain the interventions. HCW satisfaction in response to organisational changes is a dynamic construct, as illustrated in the Kübler-Ross model.¹⁶ In the ideal set-up, the interventions should be implemented in a sustained and consistent manner, in order to clearly explore the impact on HCW satisfaction and intent to stay over time. Lastly, the time period of survey administration differed between the urban site compared with the rural and remote sites. The baseline survey of the urban site was done in the year 2016, while the follow-up survey (2021 and 2022) was done 5 and 6 years after baseline. In the rural and remote sites, the baseline survey was done in 2019, with follow-up surveys done in the subsequent years without any gap years. Thus, comparison of HCW satisfaction across these sites should be done cautiously given the variability in conditions of survey administration. Due to the variation of healthcare systems and the sociocultural nuances in measuring HCW satisfaction and intention to stay, the results of this study may not be generalisable to other countries. However, these results support the need for contextualised approach to improving healthcare so as not to inadvertently aggravate inequity in resource-limited settings.

Conclusion

HCW satisfaction improved in the urban site and rural site, while HCW satisfaction declined in the remote site. Intention to stay of primary care HCWs did not significantly change before and up to 6 years after implementation of the interventions intended to strengthen primary care systems. Job involvement, longer employment and working as a nurse were associated with lower intent to stay, while enjoyment, job satisfaction and working in the urban site were associated with higher intent to stay. Strong organisational support is needed when instituting healthcare system changes to promote HCW satisfaction, especially in resource-constrained settings.

Data availability statement

Data are available upon reasonable request. The dataset is available upon request from the corresponding author.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).

Ethics approval

This study involves human participants and was approved by the University of the Philippines Manila Research Ethics Board (UPMREB: 2015-489-01) and the Philippines’ Department of Health Single Joint Research Ethics Board (SJREB: 2029-55). Verbal and written informed consent was obtained from all study participants.