The Prevalence of Compassion Fatigue and Compassion Satisfaction Among Obstetrics and Gynecology Resident Physicians in Selected Hospitals in Manila

Introduction

In recent decades, the idea of compassion fatigue (CF) has gained increasing attention. It was first mentioned in relation to healthcare, emergency and community service professionals as early as the 1990s, as compassion and empathy shown by these person-oriented professionals can be costly to one’s physical, mental and economic well-being. Exposure to patients and their traumas or distress can negatively impact a professional’s health and quality of life, which may indirectly affect the people they care for, including their families.[1,2]

CF was introduced as a state of tension and preoccupation with traumatized patients—a more “user-friendly” term for secondary traumatic stress (STS)—characterized by the desire to avoid patients and their traumas and reduced interest in bearing their suffering.[1] Distinct from BO, the main cause of CF is repeated exposure to suffering, often described as a rescue-caretaking response.[1,3]

Healthcare professionals, who are expected to deliver the highest quality of care, are among those most at risk for CF. Symptoms may affect cognitive, emotional, behavioral, spiritual, interpersonal, somatic and professional domains.[2-4] Despite these negative impacts, many professionals are motivated by a sense of gratification from helping others, a phenomenon referred to as compassion satisfaction (CS).[3,4]

In contrast to CF, CS refers to the emotional fulfillment derived from helping others and the belief that one can positively impact others through social connection. This satisfaction may serve as a coping mechanism and buffer against CF, offering strength even amid poor patient outcomes and stress—ultimately preventing BO.[2,3]

Despite the high risk, limited research has been done on the prevalence of CF among obstetricians–gynecologists (OBGYNs).[5-7] OBGYNs play an integral role in maternal and fetal care and are frequently exposed to emotionally intense scenarios. While many experiences are joyous, they may also include traumatic births and abortion care, which can result in occupational distress, fear of blame and feelings of guilt.[6-8]

The goal of CF and CS research is the early recognition of CF symptoms among healthcare providers. Exposure to secondary trauma can lead to emotional exhaustion and more serious mental health outcomes.[2,9] Thus, adequate coping strategies and interventions that promote resilience are crucial to preventing CF and related disorders, such as anxiety and depression, which may affect overall well-being and clinical productivity.[2,9,10]

The objectives of this study are to: 1) determine the prevalence of CF and CS among obstetrics and gynecologic resident physicians in selected hospitals in Manila and 2) identify factors that may contribute to CF among obstetrics and gynecologic resident physicians in selected hospitals in Manila.

Materials and Methods

Ethics

This study received ethical approval on November 11, 2024, and adhered to the National Ethical Guidelines for Health and Health-Related Research (2017 Edition). Participants were assured of their right to withdraw at any time without obligation or need to provide a reason. Full disclosure of the study’s nature and purpose was provided to allow informed decision-making, and no personally identifying information was collected to maintain anonymity. The researchers ensured that all participants understood the information given, protected their privacy and personal data, minimized potential risks, prioritized participant safety and well-being and included plans for sharing the study's benefits with the population involved.

Study Design

This study employed a quantitative, cross-sectional, descriptive-analytic design. It aims to determine the prevalence of CF and CS among OBGYN resident physicians in selected hospitals in Manila, as well as to identify factors that may contribute to CF among them.

A cross-sectional design was chosen as it allows for collection of data from a population at a single point in time. This is appropriate for assessing prevalence and associations between variables such as sociodemographic characteristics, work-related factors and levels of CF and CS.

Study Setting

The study was conducted in selected public and private tertiary hospitals located in Manila that offer accredited OBGYN residency training programs. The selected hospitals include Dr. Jose Fabella Memorial Hospital, Jose R. Reyes Memorial Medical Center, Manila Doctors Hospital, Mary Johnston Hospital, Medical Center Manila, Philippine General Hospital, Ospital ng Maynila, Tondo Medical Center, University of Santo Tomas Hospital, Metropolitan Medical Center and Our Lady of Lourdes Hospital. The recruitment process began from November 2024 to January 2025, while data collection started in November 2024 up until May 2025.

Study Population

The study population consisted of OBGYN residents currently undergoing accredited residency training in selected tertiary hospitals in Manila. These residents were eligible regardless of their year level, provided they were actively enrolled in the training program at the time of data collection and had given their consent to participate in the study.

Residents who resigned or graduated at the time of data collection, those with incomplete responses and residents that were not given questionnaires or were unavailable during data collection were excluded from the study.

Study Variables

The dependent variables are CF and CS, which are both measured using the Professional Quality of Life Scale (ProQOL) Version 5.[11] CF refers to the negative aspects of providing care to those who have experienced extreme or traumatic stressors. This includes two components – BO and STS. CS refers to the feeling of being satisfied by one’s job and from the helping itself. Each of these is measured as a continuous variable using the ProQOL 5 scoring system and can be categorized as low, average or high based on established cut-off scores.

The independent variables in our study include demographic data (age, sex, marital status and number of children), training-related data (year level, number of duty hours per week, number of hours of sleep per day and number of times being called back from off-duties) and institutional setting (private, government).

Study Size

The minimum sample size for this study was calculated using the Raosoft sample size calculator with the following parameters: a 5% margin of error, 95% confidence level, a population size of 222 obstetrics and gynecology resident physicians across selected hospitals in Manila, and an expected response distribution of 50%. These parameters generated a minimum required sample size of 141 respondents to achieve results with adequate accuracy and representativeness.

Due to practical considerations such as accessibility and availability of participants during the study period, convenience sampling was employed. Residents readily available and willing to participate were recruited until the sample size was met.

Study Procedure

1. The researchers will create an online questionnaire which will be distributed through Google Forms.

2. The link for the form will be given to Obstetrics and Gynecology resident physicians through coordination with the administration or resource person/s of the selected hospitals in Manila.

3. Data in the form of ratio variables will be gathered from the questionnaires and subjected to statistical analysis.

4. The researchers will utilize the measure of dispersion in summarizing our data. They will make use of the standard deviation to analyze it.

5. The outcomes to be measured in this study are:

   1. Compassion Fatigue (CF): Burnout (BO) and Secondary Traumatic Stress (STS)

   2. Compassion Satisfaction (CS)

   3. Factors that may affect CF and CS scores

Data Collection Tool

The Professional Quality of Life Scale, known as the ProQOL, is the most commonly used measure of the positive and negative effects of working with people who have experienced extremely stressful events.

The reliability of the three scales was calculated using both the standard methods, Cronbach's alpha and composite reliability (CR). With respect to the Australian nurses dataset, CS had α = 0.90, CR = 0.92; STS at α = 0.82, CR = 0.88; and BO at α = 0.80, CR = 0.83. With respect to the Canada nurses dataset, CS had α = 0.91, CR = 0.93; STS at α = 0.85, CR = 0.89; and BO at α = 0.75, CR = 0.79. With respect to the palliative nurses dataset, CS had α = 0.89, CR = 0.93; STS at α = 0.78, CR = 0.86; and BO at α = 0.74, CR = 0.78. Thus, each scale displayed reliability values in excess of 0.70, the recommended minimum for confirmatory studies, providing evidence supporting reliability of the ProQOL 5′s three scales across the three datasets.

There is good construct validity with over 200 published papers. There are also more than 100,000 articles on the internet. Of the 100 published research papers on CF, STS and vicarious traumatization, nearly half have utilized the ProQOL or one of its earlier versions. The three scales measure separate constructs. The CF scale is distinct. The inter-scale correlations show 2% shared variance (r = -0.23; co-σ = 5%; n=1187) with STS and 5% shared variance (r= - 0.14; co-σ = 2%; n=1187) with BO. While there is shared variance between BO and STS, the two scales measure different constructs with shared variance likely reflecting the distress that is common to both conditions. The shared variance between these two scales is 34% (r = 0.58; co-σ = 34%; n=1187). The scales both measure negative affect but are clearly different; the BO scale does not address fear while the STS scale does.

Survey Structure and Components

The self-administered survey form used in this study was divided into three main sections. The first part included informed consent, which explained the study’s objectives, confidentiality of responses, voluntary participation and the right to withdraw at any time. The second part collected information on the respondents’ sociodemographic characteristics and work-related factors, such as age, sex, marital status, year level in residency, number of working hours and hospital affiliation. The third part utilized the ProQOL.

Statistical Methods

All data were encoded and analyzed using Jamovi Version 2.6.26. Descriptive statistics for the ProQOL scores were expressed in frequencies, percentages, mean scores, standard deviations and 95% confidence intervals. Descriptive statistics such as frequencies, percentages, means, standard deviations and p-values were used to summarize the sociodemographic and work-related factors.

For inferential analysis, an independent t-test was used to compare ProQOL subscale scores between two-category variables. For variables with more than two categories, the Kruskal-Wallis test was used. A p-value of <0.05 was considered statistically significant.

Results

A total of 105 OBGYN residents from selected hospitals in Manila participated in the study. Data were collected using the ProQOLquestionnaire via Google Forms. The ProQOL 5 is validated by Beth Hudnall Stamm, PhD, designed to measure CS, BO and STS among helping professionals. The researchers used The Concise ProQOL Manual (2010) in interpreting the scale scores with the following cutoff points as described on page 29 of the manual:

• Low – scores less than or equal to 22

• Moderate – scores between 23 and 41

• High – scores greater than 42

The study assessed the prevalence of CF, subdivided into BO and STS and CS among OBGYN residents in selected hospitals and the results were summarized in Table 1.

Table 1: Prevalence of compassion fatigue and compassion satisfaction per level among OBGYN residents in selected hospitals in Manila (n=105)

STS – Secondary Traumatic Stress

The findings revealed that the majority of OBGYN residents experience moderate levels of CF (72.4% BO and 73.3% STS) and CS (68.6%). Low CS was the least common (4.8%).

Table 2: Mean scores of compassion fatigue and compassion satisfaction with standard deviations and 95% confidence intervals (CI) among OBGYN residents in selected hospitals in Manila (n=105)

The mean score for BO among the obstetrics and gynecology residents was 28.8 (SD = 7.16), with a 95% CI ranging from 27.4 to 30.2. This indicates a moderate level of BO within the group. Similarly, the mean score for STS was 29.7 (SD = 8.20), with a 95% confidence interval of 28.1 to 31.2, also reflecting a moderate level of STS.

In contrast, the mean score for CS was higher at 37.5 (SD = 6.76), with a 95% confidence interval between 36.2 and 38.8, suggesting that many residents experience moderate to high levels of satisfaction derived from their work despite the presence of stressors.

The study also explored the association between sociodemographic and work-related factors and the subscales of CF and CS. Using the ProQOL 5, certain variables were found to be significantly associated with differences in the subscale scores and these may influence the professional quality of life of residents working in a high-demand clinical environment. These associations are summarized in Table 3 for sociodemographic factors and Table 4 for work-related factors.

Table 3: Association of sociodemographic with compassion fatigue and compassion satisfaction subscales among OBGYN residents in selected hospitals in Manila (n=105)

STS – Secondary Traumatic Stress

Sex

Sex did not show a statistically significant association with any of the subscales of CF (BO: p = 0.080; STS: p = 0.077) nor CS (p = 0.113). This may partly be due to the small number of male respondents (1.9%), which limits the ability to detect sex-based differences.

Age

Age also did not show a statistically significant association with the subscales as well (BO: p = 0.294; STS: p = 0.278; CS: p = 0.373).

Residence

Most participants were from the urban areas (99, 94.3%), while only 6 (5.7%) were from rural areas. Residence also had no statistically significant association with the subscales (BO: p = 0.251, STS: p = 0.797, CS: p = 0.949).

Religion

Majority were Roman Catholic (87, 82.8%), followed by Non-Catholic Christians (16, 15.2%). There was a statistical difference noted in the STS scores (p = 0.006), but not for BO (p = 0.251) or CS (p = 0.116).

Marital Status

Single residents comprised 96 (91.4%) of the study participants, while the rest were married. All differences were statistically significant. (BO: p = 0.006, STS: p<0.001, CS: p = 0.004). Married residents had higher BO (35.0, SD = 11.1) and STS (38.2, SD = 12.1) scores and lower CS (31.3, SD = 9.84) scores compared to single residents (BO: 28.2, SD = 6.46; STS: 28.9, SD = 7.33; CS: 38.1, SD = 6.16).

Parity

Most residents were nulliparous (93, 88.6%), while the rest (11.4%) had at least one child. Significant associations were found for BO (p = 0.040) and STS (p = 0.003), but not for CS (p = 0.085). Residents with children had higher mean BO (32.8, SD = 10.4) and STS (36.1, SD = 11.3) scores compared to nulliparous residents (BO: 28.3, SD = 6.54; STS: 28.8, SD = 7.39).

Table 4: Association of work-related factors with compassion fatigue and compassion satisfaction subscales among OBGYN residents in selected hospitals in Manila (n=105)

STS – Secondary Traumatic Stress

Type of Institution

The type of institution correlates significantly with BO (p = 0.021). BO scores were higher among OBGYN residents in public hospitals (mean = 29.8, SD = 7.30) compared to those in private hospitals (mean = 26.2, SD = 6.20). There were no statistically significant differences between public and private hospitals in terms of CS (p = 0.058) and STS (p = 0.230).

Level of Training

The level of training and STS were considerably correlated (p = 0.026). STS was highest among fourth-year residents with a mean (SD) of 38.8 (5.78), followed by first-year at 38.5 (6.05), second-year at 36.9 (7.03) and lowest in third-year residents at 35.0 (8.21). There was no statistical significance with the BO (p = 0.130) and CS (p = 0.288) scores.

Work Hours per Week

BO scores increased with longer work hours, with mean values of 22.7, 26.1, 29.4 and 33.1, respectively, showing a statistically significant difference (p = 0.037). A similar trend was observed in STS scores, with means of 22.2, 26.1, 30.4 and 35.4, respectively (p = 0.013). Conversely, CS scores were highest among those working fewer hours, with mean scores of 41.5 for those working less than 60 hours, decreasing progressively to 34.6 among those working more than 100 hours per week (p = 0.041).

Sleeping Hours per Day

BO (p<0.001) was strongly correlated with the number of hours of sleep per day. BO was significantly higher among residents who reported sleeping fewer than six hours per day (mean = 29.8, SD = 6.94) than among those who slept at least six hours (mean = 23.9, SD = 6.24). Although STS scores were also higher among those with less sleep, the difference was not statistically significant (p = 0.190). On the other hand, CS was significantly greater (p = 0.002) among residents with more sleep (mean = 41.7, SD = 4.39), compared to 36.6 (6.85) in those who slept less.

Number of Hours Called in From Off-Duties per Month

In all three domains (CS: p<0.001, BO: p<0.001; STS: p<0.001), the monthly number of hours called in from off-duty was significant. The residents with highest CS (mean = 41.0, SD = 5.73) and lowest BO (mean = 24.2, SD = 5.38) and STS (mean = 24.2, SD = 5.50) scores were those who were not called in during off-duty hours. In contrast, the highest BO was reported by those who were called in two to three times per month (mean = 33.7, SD = 6.92).

Discussion

This study aimed to determine the prevalence of CF and CS among obstetrics and gynecologic resident physicians in selected hospitals in Manila, as well as to identify factors that may contribute to CF among them. Despite inadequately reaching the calculated target sample size, the findings offer valuable insights regarding the quality of life among residents in high-pressure clinical settings.

Based on the cutoffs provided in the ProQOL manual, the results revealed that the majority of OBGYN resident respondents experienced moderate levels of both BO and STS, which jointly comprise CF. Concurrently, 68.6% of respondents reported moderate CS, with 26.7% respondents achieving high CS. These results were aligned with how healthcare professionals, especially those in high-stress specialties and exposed to emotionally taxing scenarios derive personal gratification and retain a sense of professional fulfillment from their clinical roles amid stressors and balance it with their emotional exhaustion.[12]

CF was found to be particularly influenced by several personal and work-related variables. Among the contributing factors, marital status emerged as a significant one, with married resident physicians reporting higher BO and STS while also exhibiting lower levels of CS compared to their single counterparts. These findings are consistent with earlier studies indicating that additional responsibilities outside of work, especially familial ones, can compound occupational stress.[13] In relation to this, resident physicians who already have their own children also showed elevated levels of CF, which may reflect the strain of navigating parental duties alongside work obligations.

Interestingly, religious affiliation was found to significantly influence STS scores, but not BO and CS. This may suggest a possible association between personal belief systems and susceptibility to trauma-related stress, and can also be a reflection of different spiritual coping mechanisms, cultural expectations and varied perceptions of trauma.

Working for more than 100 hours per week and sleeping less than 6 hours per night were both strongly associated with higher BO and STS, consistent with previous studies emphasizing the detrimental effects of excessive workload and sleep deprivation on mental health and resilience.[14] Moreover, resident physicians frequently called in during their off duty hours reported having low CS and high CF levels. This highly emphasizes the importance of manageable workloads and how lack of recovery time is one of the main factors that drive emotional exhaustion.[13]

Institutional setting was also a critical determinant, with residents from public hospitals reporting higher BO levels compared to private institutions. This is likely due to heavier patient caseloads, limited resources and fewer systemic support, which are all commonly seen in public/government healthcare services.[12] Meanwhile, no significant difference in CS was observed by hospital type. Moreover, third-year residents showed peak BO scores, consistent with prior findings that intermediate training stages bear a disproportionate burden due to role intensification without full autonomy.[15]

The study affirms that while OBGYN residents in Manila experience moderate levels of CS, a substantial proportion still maintain moderate to high CS. This study also highlights the multifactorial nature of CF. These findings reinforce urgency of systemic interventions since BO and STS are greatly exacerbated by longer hours, less sleep and more interruptions. In the meantime, CS may be maintained by encouraging work settings that permit relaxation and recuperation. In order to better support the mental health of OBGYN residents, these findings advocate for systemic and institutional changes, such as better scheduling and duty hour limitations.

Limitations

The use of convenience sampling may limit generalizability of the findings as the sample may not be fully representative of all obstetrics and gynecology resident physicians in Manila. The reliance on self-reported data may introduce reporting bias or social desirability bias, where respondents might underreport symptoms of fatigue or overreport satisfaction. Lastly, the response rate and sample size may have affected the power of the study to detect statistically significant associations, especially if the minimum sample size was not met.

Acknowledgements

We would like to acknowledge the valuable guidance and advice provided by Eva Irene Yu–Maglonzo, MD, FPAFP, MPHEd, FPCGM, who served as a content adviser and express our gratitude to Ma. Therese B. Mallen, MD, FPOGS, FIFEPAG, for serving as a resource person during the data collection process.

Support/Affiliation

The research received no external funding. The study was conducted independently as part of the authors’ academic requirement and under the supervision of the Department of Clinical Epidemiology of the University of Santo Tomas - Faculty of Medicine and Surgery.

Conflicts of Interest

We, the authors, declare that we have no conflicts of interest to disclose related to the conduct or publication of this study.