Predictive power of the HEMPA risk assessment method for musculoskeletal disorders in nurses and caregivers: insights and implications

Abstract

Introduction

Work-related musculoskeletal disorders (WMSDs) pose a critical occupational health challenge for nurses and caregivers, with high prevalence rates impacting workforce sustainability and care quality. This study evaluates the predictive power of the HEMPA (Herramienta de Evaluación de Movilización de Pacientes) risk assessment method for WMSDs in these high-risk populations, addressing the urgent need for validated tools to guide ergonomic interventions.

Methods

This descriptive and analytical study involved 90 caregivers and nurses from various wards of a medical center in 2024. Data were collected through a three-part form supervised by researchers (demographics, body map musculoskeletal questionnaire, HEMPA tool). Logistic regression assessed HEMPA’s predictive capacity for WMSDs prevalence and its relationship with variables like age, gender, body mass index (BMI), and job characteristics. The predictive power of the HEMPA method for different body parts was evaluated using the area under the receiver operating characteristic (ROC) curve values. The study data were analyzed using Stata v.14 software, with a significance level of less than 5% for all tests (P < 0.05).

Result

In this study, 90 caregivers and nurses participated with an age range of 24–60 years and BMI of 27.15 ± 4.02 kg/m². Most of the participants (52.2%) were male, married (83.3%), and had a high school diploma (81.1%). The risk assessment of 16 different departments of the hospital was at the medium level. The highest prevalence of musculoskeletal disorders was reported in the lower back (93.3%) and neck (87.7%), while the highest intensity of pain was reported in the lower back (34.4%) and back (24.4%). The odds of suffering from musculoskeletal disorders were 0.47 times higher in the Left-Thigh (95% CI: 0.29–0.76) and Right-Thigh (95% CI: 0.29–0.76) areas compared to other body parts. Based on the ROC Curve values, the highest AUC values were for Left-Thigh (AUC = 0.79, 95% CI = 0.69–0.89) and Right-Knee (AUC = 0.76, 95% CI = 0.62–0.90), while the lowest AUC values were for Left-Ankle (AUC = 0.68, 95% CI = 0.57–0.79) and Right-Hand (AUC = 0.66, 95% CI = 0.55–0.78).

Conclusion

HEMPA effectively identifies WMSD risks, particularly in high-burden areas like the lower back and neck. This validation underscores HEMPA’s utility as a preventive tool, aligning with global efforts to reduce ergonomic hazards for caregivers and nurses. Furthermore, integrating HEMPA with digital monitoring platforms such as injury reporting software and wearable sensors for real-time risk assessment is proposed as a practical solution.

Clinical trial number

Not applicable.

Introduction

Work-related musculoskeletal disorders (WMSDs) rank among the most prevalent and debilitating occupational health challenges globally [1, 2]. According to the World Health Organization report (2022), approximately 1.71 billion people worldwide suffer from musculoskeletal disorders, with lower back pain being the leading cause of disability across 160 countries [3]. These disorders, primarily triggered by ergonomic factors such as repetitive motions, awkward postures, and heavy lifting, impose significant physical, psychosocial, and economic burdens on individuals [4, 5]. Consequently, WMSDs have emerged as a critical public health concern and a major focus of ergonomic initiatives [6–8].

Healthcare workers (HCWs), particularly nurses and caregivers, face a heightened risk of WMSDs due to multiple work conditions and occupational demands such as patient handling, maneuvering heavy equipment, and prolonged caregiving tasks [8–11]. Studies estimate the prevalence of WMSDs among HCWs to range between 43% and 78% [12, 13]. Specifically, nurses exhibit a notably higher prevalence of 79% [14], with over 30% of body areas affected in three-quarters of cases [15].

The most common anatomical sites of WMSDs among nurses include the lower back (59.5%) and neck (53%) [16]. Furthermore, among caregivers and healthcare professionals, the lower back (over 60%) and the shoulder and upper extremities (35–55%) are reported as the most prevalent sites of musculoskeletal disorders [17, 18]. Beyond individual health consequences, these disorders compromise the quality of patient care, contribute to frequent work absenteeism and early retirement, and may lead to premature workforce attrition or career changes [17, 19].

The risk factors for WMSDs in HCWs are multifactorial. Thus, individual predictor variables including gender, age, body mass index, smoking habits, exercise or other activities involving significant physical activity, medical history, and obstetric history can be engaged in the variation of WMSDs [13, 20, 21]. Moroever, organizational factors such as increased workload pressures and tight work schedules, ergonomic factors including workstation design, equipment usability, and adherence to ergonomic guidelines in clinical settings, and psychological risk factors such as job stress, time pressure, and emotional strain reported by participants particularly in high-demand units exacerbate the prevalence of WMSDs among nurses and caregivers [22, 23]. Therefore, the physical and mental health of healthcare personnel is critical for providing high-quality healthcare services to patients [24, 25].

Previous studies have found that the identification of ergonomic risk factors, the intervention and correction of ergonomic conditions, and compliance with the principles of ergonomics in the workplace have a significant effect on reducing the severity of injuries and musculoskeletal disorders, reducing medical costs, reducing psychological stress, and increasing production, job satisfaction, performance and productivity in health and nursing staff [26–30].

Several tools, including the movement and assistance of hospital patients (MAPO), patient transfer assessment instrument (PTAI) [31], and direct observation instrument for assessment of nurses’ patient transfer technique (DINO) methods [32] and Care Thermometer and the Dortmund approach have been developed to identify and assess ergonomic risk factors risk in the workplace. However, many of these tools focus narrowly on physical factors (e.g., postural analysis) and lack the comprehensiveness to address environmental, psychosocial, and organizational variables. For instance, studies employing the MAPO, PTAI and DINO methods reported weak correlations between risk scores and actual WMSDs prevalence [33–35].

In contrast, the HEMPA method, a novel and comprehensive tool, integrates 10 key components in patient handling such as patient dependency, environmental conditions, equipment availability, staff training, and risk perception to holistically evaluate ergonomic risks (Table 1) [11, 36]. By employing quantitative scoring systems, HEMPA not only addresses the limitations of prior methods but also enables targeted risk prediction for specific body regions [24].

Table 1.

Comparison between six risk assessment methods of patient handling (✓: yes; ×: No)

Valued items	MAPO	DINO	PTAI	Care	Dortmund	HEMPA
1. Specificity	✓	✓	✓	✓	✓	✓
2. Dependency Level	✓	×	×	✓	✓	✓
3. Environmental Conditions	×	×	✓	✓	✓	✓
4. Workspaces	✓	✓	✓	✓	✓	✓
5. Mechanical Aids	✓	✓	✓	✓	✓	✓
6. Postural Analysis	×	✓	✓	✓	×	✓
7. Task Outturn	×	✓	×	×	×	✓
8. Work Organization	✓	×	✓	×	×	✓
9. Training	✓	×	✓	×	×	✓
10. Risk Perception	×	×	✓	×	×	✓

WMSDs represent a pervasive and costly health challenge with profound impacts on individuals’ familial, occupational, and social well-being, often leading to chronic pain and disability [26, 37]. Certain occupational groups such as healthcare workers, flight attendants, and greenhouse farm workers exhibit disproportionately high prevalence rates, significantly undermining workforce productivity and quality of care [38]. In healthcare settings, caregivers and nurses are particularly vulnerable due to the physical demands of patient handling, yet existing research frequently overlooks the multifactorial nature of WMSDs by isolating individual risk factors rather than examining the interplay of demographic, organizational, environmental, psychological, and occupational variables.

While tools like the HEMPA method have emerged to address these gaps by holistically assessing risks and demonstrating a positive correlation between risk exposure and injury, their predictive validity remains underexplored, especially in resource-limited or developing contexts [24]. This study aimed to bridge this critical knowledge gap by rigorously evaluating potential of HEMPA approach to predict WMSDs prevalence among nurses and caregivers, with a specific emphasis on the multifactorial interactions underlying these disorders and its utility in identifying at-risk populations and informing targeted interventions through comprehensive ergonomic risk assessment. Ultimately, this study was conducted based on the main hypothesis that the HEMPA method effectively predicts the prevalence of WMSDs in different body regions among nurses and caregivers with high statistical power.

The implementation of this study highlights the potential to integrate HEMPA into national monitoring programs for the systematic identification of high-risk sectors, the design of targeted managerial and engineering training and control programs based on the prioritization of high-risk areas, and the development of evidence-based ergonomic intervention guidelines to reduce WMSDs among HCWs.

Methods

Study design

This cross-sectional analytical descriptive study was conducted on 90 caregivers and nurses across various departments of a hospital in 2024. Informed consent was secured from all participants prior to inclusion and the questionnaires were self-reported by participants. Data collection was carried out confidentially and anonymously, with occupational health and ergonomics researchers present on-site to address conceptual inquiries. However, no interference occurred in the content of participants’ responses.

The study included participants aged 24–60 years with ≥ 12 months of cumulative experience in direct patient-handling tasks (e.g., lifting, transferring), irrespective of workplace continuity. On the other hand, Participants were excluded from the study with history of systemic diseases affecting musculoskeletal system (e.g., diabetes mellitus with complications, chronic kidney disease, autoimmune disorders, osteoarthritis, rheumatoid arthritis and osteoporosis) or acute traumatic injuries from non-occupational accidents (e.g., fractures from falls or car accidents). These criteria aimed to isolate WMSDs while controlling for age-related degenerative conditions and non-occupational confounders.

Sampling method & sample size calculation

The sample size determined about 90 using the study of Yousefi Seyf [24] and prevalence of about 20% for WMSDs in the left and right wrist, accuracy of 5%, and also based on the sample size Eq. (1) for limited communities (150 people in this study).

Where Z = Z value (1.96 for 95% confidence level), p = 0.2, q = 0.8, d = 0.05.

After determining the sample size, a disproportionate stratified random sampling was conducted. For this purpose, a comprehensive list of personnel (nurses and caregivers responsible for patient handling) was first compiled based on each hospital department. Each case was then assigned a unique identification number. For each department, proportional to the number of personnel, a random number generator (Excel RAND function) was used to assign random sampling values to all eligible personnel. Participants were selected in ascending order of these values until the quota for each specific department was reached. Ultimately, 90 nurses and caregivers from 16 departments participated in the study (Table 5). All stages of the participant recruitment process were carried out in coordination with department managers and through the organization of briefing sessions and data collection spanned six months (February–July 2024).

Data collection techniques

In this study, a three-part form was utilized for data collection, which included demographic information, a body map questionnaire, and the HEMPA tool. The body map questionnaire was employed to determine the prevalence of musculoskeletal disorders. The different departments of the hospital where the caregivers and nurses worked and where there was a need to handle the patient were investigated using the HEMPA questionnaire. Moreover, environmental factors such as noise, lighting, and heat stress were measured to confirm parts of the HEMPA questionnaire based on objective assessment.

Demographic information

This demographic tool contained job-related topics and individual characteristics such as age, weight, height, marital status, and job authentication information such as job history and working hours per week and month.

Body map questionnaire

The standardized Body Map tool, developed by McGill, was designed to identify the location and severity of musculoskeletal discomfort and pain [39]. This collaborative research tool, recognized for its ease of access, face validity, and applicability in self-reported research, has been widely adopted as one of the most utilized assessment methods and has been successfully employed to investigate workplace health issues across diverse occupational settings [24, 40, 41]. In time-constrained healthcare environments, where HCWs face heavy workloads and limited availability for research participation, this standardized tool accelerates data collection and improves participation rates. The original McGill version has undergone revisions in multiple studies due to limitations such as inadequate alignment of anatomical boundaries with actual pain patterns and the absence of condition-specific clinical features [42]. In the Persian adaptation of this tool, the body is divided into 19 distinct regions. Participants mark pain locations on a body diagram and report pain intensity using a 5-point scale, ranging from no pain (1) to severe pain (5) [43]. Pain-free regions are classified as absence of musculoskeletal disorders, with higher scores indicating greater discomfort. The content and structural validity of this questionnaire have been confirmed in prior studies (intra-rater agreement: 0.94 ± 0.01; r = 0.64, P < 0.01) [44, 45].

Environmental factor assessment

To measure the noise, the dimensions of each room were divided into squares of 2 × 2 m², and then the sound level (dB) was measured in the center of each square (station) using a sound level meter at the hearing height of the people. Noise levels were measured using a TES-1358 C sound meter (TES Electrical Electronic Corp., Taiwan); room dimensions were measured with a Leica DISTO D2 laser meter (Leica Geosystems, Switzerland). Then, Eq. (2) was used to calculate the average sound pressure level (LP or SPL) in the room.

Equation (2):

where LP: Average sound pressure level, n: Number of measuring stations, LPi: Sound pressure level at each station.

Humidity and temperature were measured using heat index WBGT meter (Lutron WBGT-2010SD WBGT Meter, Taiwan) at one point of the room after being present for 10 min to achieve temperature and humidity balance. Illumination in terms of Lux was also measured in each room using the digital illuminance meter (TES-1339R, TES Electrical Electronic Corp, Taiwan) in the center of each 2 × 2 m² square (station) at the height of the work station. Occupational health and ergonomics experts conducted all measurements and verified the accuracy of the findings. These experts had completed two academic units (32 h) in patient care ergonomics at the university and received appropriate training in measurement techniques. They had also practiced environmental factor measurement methods in laboratory settings. This demonstrates a commitment to the precision and reliability of the collected data. This information was collected for decision making in the subscales of the HEMPA questionnaire. The workplace assessment was conducted in accordance with NIOSH guidelines for ergonomic factors and ISO 9241 (workplace ergonomics) standards. Minimum ambient lighting levels were applied based on ISO 8995 (500 lx for clinical areas). All measurements were verified to comply with Iran’s Occupational Exposure Limits (OELs) guidelines (5th edition, 2021) [46].

HEMPA risk assessment method

HEMPA is a method for assessing the patient’s handling conditions, which was developed and used by Villarroya et al. in hospitals in 2016, and the reliability and validity of this tool was confirmed in 2017 [11]. In addition, based on Yousefi Seyf’s study, the validity and reliability of the Persian version of the HEMPA tool was confirmed in 2023 [24]. In the HEMPA method, 10 of the most critical items in patient handling are accurately evaluated and the risks related to patient care are correctly determined.

In the HEMPA method, 10 items such as dependency level, environmental conditions, working spaces, suitable equipment and technology as minor aids, suitable equipment and technology as major aids, handling implementation and situation analysis, treatment result, work organization, training, and risk understanding were accurately and the risks related to patient care were determined (Table 2).

Table 2.

HEMPA method outline

HEMPA item	Explanation
Dependency Level	Combines patient mobility, cooperation, and dependency
Environmental Conditions	Evaluates noise, brightness, humidity, and temperature
Workspaces	Assesses bathroom accessibility, toilet and room space
Minor Aids	Reviews availability of small aids (e.g., sliding sheets, walkers)
Major Aids	Assesses large equipment (e.g., patient lifts, adjustable beds)
Transfer Execution	Analyzes 10 manual handling methods (e.g., moving patient to chair)
Handling Outcome	Evaluates patient comfort
Work Organization	Includes patient-to-nurse ratio, rest breaks, night shifts, and peer support
Training	Examines staff training on handling risks, practical techniques, and device use
Risk Perception	Assesses caregiver burden via four questions (e.g., physical harm, preplanning)

The items of the HEMAP method are as follows:

1. Dependency level.

This scale, which is obtained by combining mobility, cooperation, and patient dependency, has a maximum score of 3. Mobility is divided into five levels (score between 0.6 and 3: fully bedridden patients to independent patients) and cooperation and patients’ dependency into three levels (score between 1 and 3: patients who are not cooperative to cooperate). The total score of this item for each patient was obtained by dividing the scores of mobility and cooperation by 2. The final score of this item is calculated for all patients according to Eq. 3:

Equation 3:

• b)Environmental conditions.

This scale, which has a maximum score of 1, measures noise in 24 h (standard level of 40 dB during the day between 7:00 and 11:00 p.m. and 30 dB at night between 11:00 p.m. and 7:00 a.m.), brightness (minimum standard 500 lx), humidity (standard 30–70%), and temperature (standard between 14 and 25 degrees Celsius). If each subscale is met, a score of 0.25 is given. The total score of this item is obtained by adding the scores of all sub-items for each room. The final score for all rooms was calculated according to Eq. 4.

Equation 4:

• c)Workspaces.

This scale, which examines the accessibility and physical space of the bathroom, toilet features, and the possibility of adjusting the bed to care for patients and rooms, has a maximum score of 5. Most of the described features depend on the degree of compliance with the regulations governing architectural barriers. Therefore, they are as follows: bathroom (unobstructed access to the bathroom, door width of at least 85 cm and enough space for proper movement of mechanical aids), toilet (toilet cup height of at least 50 cm and the presence of a side support bar next to the toilet); suitable work space for moving a wheelchair), and room (space between beds at least 90 cm; free space from the foot of the bed to the wall at least 120 cm); The possibility of mechanical adjustment of the beds, both in height and in the inclination of the head of the bed. Bathrooms, toilets, and a room score between 0: inappropriate and 0.625: appropriate. The possibility of mechanically adjusting the height and inclination of the bed head is checked with a score between 0: if it is inappropriate and 1.25: if it is appropriate. The score of this item is obtained by summing the scores of the sub-items for each room. To obtain the total score of this scale, the scores of all rooms were added and then divided by the total number of analyzed rooms. The final score for all rooms in each section was calculated using Eq. 5.

Equation 5:

• d)Proper equipment and technology (minor aids).

This scale has a maximum score of 5, and 1.25 points are given to any type of device or auxiliary device. Based on this this scale, available and small auxiliary equipment is examined for lifting; ambulation or handling patients, which include a sliding sheet, transfer platform, rotating or turntable disk, and a walker or standing lift. Any type of assistive device is rated and scored only if it has already met all the requirements. Based on this, the auxiliary devices in the considered unit are available in sufficient quantity for the specific movement and are under proper maintenance conditions. The final score of this scale is calculated using the following Eq. 6.

Equation 6:

• e)Proper equipment and technology (major aids).

This scale examines the availability of equipment to assist in lifting or transferring the patient (mechanical lifting devices, etc.) similar to the other analyses and has a maximum score of 5. A maximum of 1.25 points are awarded for each type of device and auxiliary device. Accordingly, it reviews the availability of equipment designed to assist in lifting or moving the patient, including patient lifts, wheelchairs, height-adjustable beds, and height-adjustable stretchers. Any assistive device will only be rated if it meets all the requirements. The equipment is also available in the unit and in sufficient quantity. They are adequate for the specific task and are in good working order. The final score for this item is calculated using Eq. 7.

Equation 7:

• f)Transfer execution and postural analysis.

This scale analyzes the 10 main manual methods of patient handling and has a maximum score of 4. These methods include: lifting the patient to a sitting position, moving the patient to the head of the bed, moving the patient to one side of the bed, raising the patient’s legs, tilting the head of the bed, moving the patient’s pelvis, placing equipment, transferring the patient from the bed to the bed, transfer from the bed to the chair, and lifting the patient from a sitting position to a standing position. A total of 0.4 points will be awarded for each task that is performed in an acceptable manner. To be considered “acceptable” means that the task is performed without adopting any inappropriate posture. Otherwise, no points will be awarded. The final score of this item is obtained by adding up the scores of 10 postures for each patient. The final score of this item was calculated for all patients using Eq. 8.

Equation 8:

• g)Handling outcome.

This scale has a maximum score of 2 and four subscales, each of which is assigned a score of 0.5. This scale is intended to estimate each of the reflected situations, the purpose of which is to check the correct implementation of patient transport. If the handling technique used does not cause any pain or discomfort to the patients, a score of 0.5 is assigned to this item. In addition, if the handling technique does not cause fear or uncertainty in patients, a score of 0.5 is assigned to this item. If the handling is not done quickly and hastily, a score of 0.5 is given, and if the patient is in a good position at the end of the handling, a score of 0.5 is given. The final score of this scale is obtained by summing the scores of four sub-items for each patient. The final score for all patients was calculated using Eq. 9:

Equation 9:

• h)Work organization.

This scale has a maximum score of 4, which includes considering the speed of work and rest, the ratio of patient to nurse or caregiver, night shift services, and peer support to care for patients. If the patient-to-nurse or caregiver ratio and peer support are sufficient to handle the patients, a score of 0.5 is assigned to each, and for work-rest speed (with two sub-items: patient transfer is done without time pressure; periodic breaks for rest is arranged) and night shift service (with two sub-items: no night work; there is at least one rest day until the nurse or caregiver returns to work), a score of 0.25 is considered for each sub-category. The final score is obtained by adding the scores of the four subcategories for each section. Ultimately, the overall score is obtained by dividing the sum of the scores by the total number of sections analyzed.

• i)Training.

This scale, which has a maximum score of 2, examines specific training in manual patient care using four items, which is a key aspect of manual patient care. These items include information about the risks associated with manual patient handling in the workplace, theoretical and practical training on manual patient handling and maintenance to at least 75% of the department’s staff, and practical training in the use of mechanical assistive devices provided in the last two years. The evaluation of the validity of the training is based on its effectiveness in reducing accidents, and if it is realized, a score of 0.5 will be awarded to each subcategory. The final score is obtained by adding the scores of the four subcategories for each section. Ultimately, the total score is obtained by dividing the total score by number of analyzed sections.

• j)Risk perception.

This scale with four subcategories has a maximum score of 1 and is determined based on some psychosocial factors by asking caregivers if there is physical or mental burden. The four sub-items are as follows: Do the working positions adopted during patient handling cause any harm to your health? Is the patients’ transfer preplanned? Do you think the patients treated are light or relatively heavy? Is the movement of patients not continuous or is it far from the work shift? If each subcategory is suitable, it will be given a score of 0.25. The final score was obtained by adding the scores of the four subcategories for each nurse. The final score for all participants in each section was calculated according to Eq. 10.

Equation 9:

Total HEMPA score

To obtain the total HEMPA score of each section and the level of risk related to each evaluated unit or service, the total HEMPA scores of all 10 items were summed. The obtained score is then divided into the following three risk levels according to Table 3. The maximum score of the total HEMPA score obtained is 30. The HEMPA risk categories (Green/Yellow/Red) were confirmed based on Villarroya et al. [11], with thresholds validated in Persian contexts [24].

Table 3.

Total HEMPA score ranges and corresponding risk levels

Risk levels	Score range	Interpretation
1 Green	20.01-30	The risk of musculoskeletal disorders is acceptable.
2 Yellow	10.01-20	The risk of musculoskeletal disorders is moderate.
3 Red	0.8–10	The risk of musculoskeletal disorders is unacceptable.

Data analyses

The independent variables, including environmental factors and workplace ergonomic conditions analyzed through the HEMPA method, as well as job-related characteristics such as task type and working hours, and individual factors including age, gender, and BMI, were investigated as influencing factors on the prevalence of WMSDs, which served as the dependent variable. The Kolmogorov-Smirnov test was used to check the normal distribution of the obtained data. Descriptive statistics (mean and standard deviation as well as median and interquartile range for quantitative variables and frequency and percentage for qualitative variables) were used to determine descriptive objectives. Logistic regression was used to investigate the HEMPA tool’s ability to predict the prevalence of musculoskeletal disorders and to investigate the multivariate relationship between the outcome and different variables. The predictive power of the HEMPA technique in different parts of the body was reported based on the Area under the ROC Curve (AUC) values. Therefore, the closer the ROC curve is to the upper left corner of the graph, the higher the accuracy of the test, because in the upper left corner, sensitivity = 1 and false positive rate = 0 (specificity = 1). Therefore, the ideal ROC curve has AUC = 1.0. The data analysis of the study was performed using Stata version 14 software, considering a significance level of less than 5% (P < 0.05) for all tests.

Result

This descriptive-analytical study was conducted in 16 different departments of a medical teaching hospital in 2024. Given the research team’s close oversight of the participation process and questionnaire administration, all 90 participants voluntarily completed the questionnaires with no dropouts (Response rate = 100%). The mean age and BMI of caregivers and nurses were determined to be 24–60 years and 27.15 ± 4.02, respectively. Most participants were male (52.2%), married (83.3%), and held a high school diploma (81.1%). Additional information about the participants is shown in Table 4.

Table 4.

Socio-demographic characteristics of the participants

Variables		N (%)
Age *	Year	37.92 (7.77)
Gender	Women	43 (47.8)
	Men	47 (52.2)
Marital Status	Single	14 (15.6)
	Married	75 (83.3)
	Divorced or deceased spouse	1 (1.1)
Education Level	High school	73 (81.1)
	Diploma	7 (7.8)
	Bachelor	9 (10.0)
	Master	1 (1.1)
Weight *	Kg	75.82 (11.92)
Height *	cm	167.12 (7.97)
Body mass index (kg/m²)	Normal weight	31 (34.4)
	Pre-obesity	40 (44.4)
	Obesity class I	14 (15.6)
	Obesity class II	4 (4.4)
	Obesity class III	1 (1.1)
Work experience **	Year	12 (7)
Surgical history	Yes	20 (22.2)
	No	70 (77.8)
Work rotation	At rotation	79 (87.8)
	On day	9 (10.0)
	At night	2 (2.2)

^* Mean (standard deviation); ^** Median (interquartile range)

After assessing the risk of different departments of the hospital using the HEMPA method, all departments of the hospital and the final score of the integration of different departments of the hospital were classified at the medium risk level (total HEMPA score of 10.01 to 20). The highest and lowest risk score were identified for intensive care unit (ICU) departments (14.65 ± 0.01) and oral and maxillofacial surgery department (19.05 ± 0.01), respectively (Table 5).

Table 5.

HEMPA scores by hospital ward

Ward	N	%	HEMPA score
			Mean ± SD	Level of risk
Gastrointestinal	6	6.7	16.35 ± 0.01	moderate
psychosomatics	3	3.3	17.45 ± 0.01	moderate
Maternity	3	3.3	16.35 ± 0.01	moderate
Oncology	4	4.4	16.35 ± 0.01	moderate
Oral and Maxillofacial Surgery	4	4.4	19.05 ± 0.01	moderate
Hemodialysis	6	6.7	16.85 ± 0.01	moderate
Hematology	15	16.7	16.85 ± 0.01	moderate
Cardiovascular	3	3.3	16.35 ± 0.01	moderate
Nuclear medicine	2	2.2	18.65 ± 0.01	moderate
Endoscopy	3	3.3	16.45 ± 0.01	moderate
CCU	3	3.3	14.75 ± 0.01	moderate
ENT	4	4.4	18.55 ± 0.01	moderate
Room Operation	9	10.0	17.50 ± 0.01	moderate
ICU	11	12.2	14.65 ± 0.01	moderate
Emergency	10	11.1	18.85 ± 0.01	moderate
Endocrine	4	4.4	16.35 ± 0.01	moderate
Total	90	100	16.90 ± 0.01	moderate

Most caregivers and nurses suffered from pain in one or more musculoskeletal areas in the past year (Fig. 1; Table 6). Moreover, the highest prevalence of musculoskeletal disorders was reported in various body parts, including the lower back (93.3%), neck (87.7%), right knee (85.5%), back (84.4%), and left knee (83.3%) (Fig. 1). Additionally, the highest intensity of pain was reported in the lower back (34.4%), back (24.4%), and neck (22.2%), respectively.

Fig. 1.

Prevalence of musculoskeletal disorders during the last year from the body map questionnaire

Table 6.

Pain intensity during the last year from the body map questionnaire

Body part	Pain level, N (%)
	No pain (1)	Low (2)	Moderate (3)	High (4)	Very high (5)
Neck	11 (12.2)	18 (20.0)	31 (34.4)	10 (11.1)	20 (22.2)
Right-Shoulder	21 (23.3)	24 (26.7)	20 (22.2)	11 (12.2)	14 (15.6)
Right-Forearm	37 (41.1)	27 (30.0)	18 (20.0)	5 (5.6)	3 (3.3)
Right-Wrist	44 (48.9)	21 (23.3)	17 (18.9)	5 (5.6)	3 (3.3)
Right-Hand	48 (53.3)	15 (16.7)	12 (13.3)	7 (7.8)	8 (8.9)
Right-Thigh	22 (24.4)	23 (25.6)	14 (15.6)	21 (23.3)	10 (11.1)
Right-Knee	13 (14.4)	17 (18.9)	22 (24.4)	19 (21.1)	19 (21.1)
Right-Calf	36 (40.0)	18 (20.0)	15 (16.7)	12 (13.3)	9 (10.0)
Right-Ankle	34 (37.8)	15 (16.7)	16 (17.8)	7 (7.8)	18 (20.0)
Back	14 (15.6)	17 (18.9)	24 (26.7)	13 (14.4)	22 (24.4)
Left-Shoulder	20 (22.2)	24 (26.7)	23 (25.6)	9 (10.0)	14 (15.6)
Left-Forearm	37 (41.1)	29 (32.2)	18 (20.0)	4 (4.4)	2 (2.2)
Left-Wrist	47 (52.2)	16 (17.8)	16 (17.8)	7 (7.8)	4 (4.4)
Left-Hand	50 (55.6)	8 (8.9)	19 (21.1)	10 (11.1)	3 (3.3)
Lower Back	6 (6.7)	3 (3.3)	19 (21.1)	31 (34.4)	31 (34.4)
Left-Thigh	22 (24.4)	23 (25.6)	14 (15.6)	21 (23.3)	10 (11.1)
Left-Knee	15 (16.7)	17 (18.9)	20 (22.2)	20 (22.2)	18 (20.0)
Left-Calf	34 (37.8)	21 (23.3)	12 (13.3)	11 (12.2)	12 (13.3)
Left-Ankle	43 (47.8)	15 (16.7)	9 (10.0)	4 (4.4)	19 (21.1)

The prevalence of WMSDs among caregivers and nurses increased with lower HEMPA scores. Hence, nurses with a lower HEMPA score, possibly indicating a less ergonomic work environment, are more likely to develop WMSDs. According to the results of logistic regression, with an increase of one point in the HEMPA score, decreases the highest chance of suffering from musculoskeletal disorders by 53% in the left thigh (AOR = 0.47; 95% CI: 0.29–0.76), by 53% in the right thigh (AOR = 0.47; 95% CI: 0.29–0.76), by 49% in the right knee (AOR = 0.51; 95% CI: 0.29–0.90), and by 42% in the left hand (AOR = 0.58; 95% CI: 0.39–0.85). Additionally, reducing the chance of WMSDs in other areas by increasing one HEMPA score has also been significant (Table 7).

Table 7.

Association between the HEMPA score and WMSDs in nurses and caregivers

Outcomes (Body part)	Unadjusted model			Adjusted model
	Odds ratio	95% CI	P-value	Odds Ratio	95% CI	P-value
Neck	0.96	0.59–1.56	0.888	1.04	0.63–1.74	0.856
Right-Shoulder	0.67	0.45–1.01	0.057	0.70	0.46–1.05	0.092
Right-Forearm	0.67	0.47–0.95	0.026*	0.69	0.48–0.99	0.049*
Right-Wrist	0.59	0.41–0.85	0.005**	0.64	0.44–0.94	0.024*
Right-Hand	0.71	0.50–0.99	0.046*	0.65	0.45–0.95	0.026*
Right-Thigh	0.47	0.30–0.75	0.001**	0.47	0.29–0.76	0.002**
Right-Knee	0.50	0.30–0.86	0.012**	0.51	0.29–0.90	0.020*
Right-Calf	0.60	0.42–0.87	0.007**	0.61	0.42–0.91	0.015*
Right-Ankle	0.73	0.52–1.04	0.082	0.75	0.52–1.08	0.128
Back	0.70	0.44–1.11	0.136	0.72	0.44–1.17	0.187
Left-Shoulder	0.64	0.42–0.97	0.040*	0.65	0.42–1.03	0.017*
Left-Forearm	0.68	0.48–0.96	0.031*	0.73	0.50–1.08	0.121
Left-Wrist	0.72	0.52–1.01	0.058	0.81	0.56–1.17	0.273
Left-Hand	0.60	0.42–0.86	0.006**	0.58	0.39–0.85	0.006**
Lower Back	1.10	0.59–2.08	0.749	1.35	0.54–3.37	0.517
Left-Thigh	0.47	0.30–0.75	0.001**	0.47	0.29–0.76	0.002**
Left-Knee	0.63	0.40–1.01	0.057	0.63	0.38–1.05	0.077
Left-Calf	0.63	0.44–0.91	0.013*	0.64	0.43–0.95	0.027*
Left-Ankle	0.65	0.46–0.92	0.017*	0.66	0.46–0.96	0.028*

Adjusted for age, gender, BMI (continuous), work experience, and surgical history; *Significance level < 0.05; **Significance level < 0.01

Based on the area under the ROC Curve (AUC) values, the highest AUC corresponds to the left thigh (AUC = 0.79, 95% CI = 0.69–0.89) and the right knee (AUC = 0.76, 95% CI = 0.62–0.90), respectively. These findings indicate that HEMPA more accurately predicted WMSDs in these body areas. Additionally, the lowest AUC is shown for the left ankle (AUC = 0.68, 95% CI = 0.57–0.79) and the right hand (AUC = 0.66, 95% CI = 0.55–0.78), respectively, indicating that HEMPA less accurately predicted WMSDs in these areas. In general, the AUC for HEMPA is greater than 0.60 in all body parts, illustrating that the HEMPA technique is a reliable predictive tool for WMSDs across all areas (Figs. 2 and 3).

Fig. 2.

ROC curves for the six right body areas

Fig. 3.

ROC curves for the five left body areas

Discussion

The identification and utilization of comprehensive tools for hazard identification and ergonomic risk assessment are critical for mitigating WMSDs. This study evaluated the effectiveness of the HEMPA tool in predicting and assessing risks associated with WMSDs during patient handling tasks.

The findings underscore the importance of considering demographic and occupational factors, such as age, gender, BMI, work experience, and shift patterns, in understanding the prevalence and severity of WMSDs among HCWs. Notably, over 20% of participants were classified as overweight, a known risk factor for WMSDs, and shift work was identified as a significant variable influencing musculoskeletal health, which according to previous studies is one of the main risk factors for developing musculoskeletal disorders caused by work [47].

The distribution of shift work among participants highlights the importance of considering shift work in relation to musculoskeletal health. However, variations in job roles and environments such as the high physical demands in ICUs versus the controlled tasks in surgical units may influence risk exposure [48]. Future studies should stratify analyses by department-specific workflows to refine interventions. Understanding these characteristics is necessary to assess the prevalence of musculoskeletal disorders and design appropriate interventions. Luan et al. determined that factors such as age, history of musculoskeletal disease, anxiety, and absenteeism from work can play an important role in developing WMSDs [4].

The HEMPA tool demonstrated its utility in assessing WMSDs risks across different hospital departments [11]. The ICU was identified as the highest-risk department, while the Oral and Maxillofacial Surgery department had the lowest risk level. The disparity in risk can be attributed to the high volume of physical activities in units such as the ICU, including patient transfers, non-neutral body postures, psychological stress, and prolonged shifts [49]. Compared to traditional tools like MAPO and DINO, which focus narrowly on physical risks in prior studies [36], HEMPA’s integration of psychosocial and organizational factors improved predictive accuracy [11]. In contrast, static and controlled tasks, the use of ergonomic equipment, and regular rest periods in clinical settings such as maxillofacial surgery may mitigate these risks [50]. These variations highlight the need for tailored risk management strategies in high-risk areas. The HEMPA tool’s multidimensional approach allows for a comprehensive evaluation of risk factors, enabling targeted interventions to reduce WMSDs [36]. For instance, preventive measures such as ergonomic training, rehabilitation programs, and workload adjustments could be prioritized in departments with elevated risk levels [24].

The study revealed that musculoskeletal pain was most prevalent in the lower back, neck, and knees, consistent with previous research identifying these areas as particularly vulnerable among HCWs [24, 51]. These areas of the body are among the vulnerable parts in handling patient, which can be caused by repetitive and high-risk activities [52]. Biomechanical studies suggest that prolonged bending during patient transfers and static neck postures during documentation contribute to lumbar and cervical strain [53, 54]. However, differences in pain distribution across studies may be attributed to variations in study populations, methodologies, and anatomical factors [55]. These findings emphasize the need for context-specific interventions to address WMSDs effectively [51, 56].

This research evaluated the predictive capacity of the HEMPA technique, a novel multidimensional approach developed to address gaps in traditional patient-handling risk assessment tools. Derived from five prior methodologies [11, 36], HEMPA integrates physical, environmental, and organizational factors to comprehensively evaluate WMSD risks, a significant advancement over earlier tools that focused narrowly on biomechanical factors. Despite its potential, limited research has utilized HEMPA for predicting patient transport-related WMSDs [24], underscoring the novelty of this investigation.

The findings demonstrate HEMPA’s strong predictive capability of WMSDs and a one-point increase in HEMPA scores correlated with substantial reductions in WMSD likelihood 53% for both thighs, 49% for the right knee, and 42% for the left hand. These findings align with Bispo et al.’s work linking poor postures to lower limb pain and emphasize the value of HEMPA’s holistic scoring system [47]. Notably, the analysis controlled for key variables such as age, BMI, and surgical history, yet unmeasured confounders such as individual ergonomic habits or genetic factors may partially influence outcomes. The tool consistently identified high-risk anatomical regions, including the forearms, hands, thighs, and lower back, areas vulnerable to repetitive strain during patient transfers. This pattern reinforces the critical need for comprehensive risk assessments in healthcare settings. By quantifying interactions between modifiable risks (e.g., equipment availability) and fixed factors (e.g., patient dependency), HEMPA enables targeted interventions to mitigate WMSDs among caregivers [36].

The area under the ROC curve (AUC), a metric used to evaluate the performance of binary classification models, demonstrated that the HEMPA tool exhibits relatively high accuracy in detecting WMSDs among patients and nurses, particularly in the left thigh (AUC = 0.79) and right knee (AUC = 0.76). However, while these values approach the clinically acceptable threshold of 0.8, they fail to surpass this benchmark, limiting their utility as robust indicators for broad clinical applications.

For context, AUC values of 0.7–0.8 are considered moderate in diagnostic research, whereas tools like the revised NIOSH lifting equation achieve AUC more than 0.85 in industrial settings [57]. In contrast, the tool’s weaker performance in the left ankle (AUC = 0.68) and right hand (AUC = 0.66) may stem from the complexity of multi-dimensional movements or insufficient training data specific to these anatomical regions. Furthermore, although HEMPA demonstrates localized accuracy in certain areas, its overall score (AUC < 0.8) reflects a significant limitation in comprehensive WMSDs prediction. Notably, aggregating data from diverse body regions into a single composite score risks diluting the predictive power of stronger-performing areas (e.g., the left thigh). These findings align with the study by Çorbacıoğlu and Aksel, which underscores that AUC values below 0.8 hold limited diagnostic value in clinical contexts [58].

Consequently, while HEMPA may serve as an effective preliminary screening tool for specific regions, its broader application necessitates algorithmic refinements such as differential weighting of specific body regions or integration with complementary metrics (e.g., biomechanical or psychosocial factors) to achieve accuracy beyond its current level. These strengths and weaknesses highlight the necessity of interpreting results with caution and contextual awareness regarding the tool’s practical implementation.

Study strengths and limitations

This study utilized the HEMPA comprehensive tool to evaluate WMSDs risks among nurses and caregivers. Results confirmed that validated assessment tools are crucial for diagnosing and preventing WMSDs in hospitals. Unlike traditional methods (e.g., MAPO, DINO), HEMPA uniquely combines environmental, psychosocial, and organizational factors, offering both risk identification and actionable mitigation strategies. This comprehensive approach excels in detecting hidden risks in complex healthcare settings.

However, like other studies, this study may have limitations. Caregivers and nurses may experience multiple tasks on the job that can serve as facilitators of WMSDs. The cross-sectional design limits causal inference and a longitudinal follow-up will assess HEMPA’s predictive validity over time. Moreover, conducting more longitudinal studies and compare with data from other convergence research using new methods of artificial intelligence such as machine learning can discover complex relationships between the factors that cause WMSDs and more accurate predictions in different healthcare settings such as hospitals with more workers [59]. For instance, machine learning models could analyze interactions between HEMPA’s 10 domains to prioritize high-impact interventions.

On the other hand, this study utilized the Body Map questionnaire to conduct a site-specific and localized analysis of musculoskeletal disorders, designed to address the unique biomechanical demands of caregiving tasks. However, the standardized Nordic Musculoskeletal Questionnaire (NMQ), as a validated tool, may influence the direct comparability of prevalence rates and risk factors due to differences in design compared to the Body Map. To address this limitation, future studies are recommended to enhance epidemiological alignment by comparing or integrating both the NMQ and Body Map tools. For policy implementation, integrating HEMPA into national occupational safety standards with funding linked to risk reduction could enhance findings’ generalizability and strengthen preventive strategies.

Conclusion

The study revealed persistently high WMSD prevalence among caregivers and nurses, with the lower back, neck, and back exhibiting the greatest severity. HEMPA demonstrated strong predictive accuracy for assessing WMSD risk levels. Prioritizing ergonomic practices such as patient-handling protocols, HEMPA adoption, and targeted training is critical for mitigating musculoskeletal disorders. Enhanced departmental analyses, policy implementation, and ergonomic training are essential to improving workplace safety and service quality. Future research should validate HEMPA in larger, multi-center cohorts, compare it with biomechanical sensors or machine learning models, and explore integration with digital surveillance platforms (e.g., injury reporting software) to optimize risk management.

Author contributions

SVE and ASS conceptualized the study. Data collection was supervised by ASS, FA and FD and analysis was conducted by NI. The manuscript was conceptualized by SVE, ASS, and AA and drafted by AA and SVE. All authors contributed to revising the manuscript and approval of the final version. All authors read and approved the final manuscript.

Funding

This study was financially supported by the Shahid Beheshti University of Medical Sciences, Tehran, Iran (grant number: 43005800).

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Approval for the study was obtained from the Shahid Beheshti University of Medical Sciences Research Ethics Committee (Ethic No. IR.SBMU.PHNS.REC.1402.099). Each participant entered the research after completing the written consent form to conduct this research. an informed consent form was obtained from all participants. All participants signed the informed consent form and were explained the experimental procedure prior to the study onset. Participants names were then masked using a numerical code after data collection to maintain confidentiality. All experiments were performed in accordance with relevant guidelines and regulations (such as the Declaration of Helsinki).

Consent for publication

Not applicable.

Disclosure

No potential conflicts of interest were reported by the authors.

Competing interests

The authors declare no competing interests.