Postoperative rehabilitation training adherence and influencing factors in adults with traumatic fractures in China: a cross-sectional study

Abstract

Objective

To explore rehabilitation training adherence as well as its related influencing factors among adult patients with traumatic limb fractures to provide a basis for clinical intervention strategies.

Design

Cross-sectional study.

Setting

Department of Rehabilitation Medicine.

Participants

Adults within 1 year postoperative for traumatic limb fractures and who underwent rehabilitation training were included in this study.

Interventions

Not applicable.

Main outcome measures

Patient information was collected using a self-designed general information form. Rehabilitation training adherence was evaluated using the Exercise Adherence Rating Scale (0/64, higher values=higher adherence). Barriers to and facilitators of rehabilitation were assessed using sociodemographic data, ability to perform daily life (ADL), General Self-Efficacy Scale (GSES) and Visual Analogue Scale. Pearson’s (r) and determination (R²) correlation coefficients were calculated to determine the strength of associations between variables. Potential correlates were explored using multiple linear regression analysis.

Results

Between December 2022 and December 2023, 170 subjects were recruited. The age of the participants ranged from 18 to 93 years, with a mean age of 41.5 years (SD 15.6). The mean Exercise Adherence score was 30.31 (SD 5.8). Adherence to postoperative rehabilitation training tended to be lower in participants of older age (−1.30; 95% CI −2.14 to −0.46; p=0.003), and those who had higher pain scores on the movement (moderate pain (−3.25; 95% CI −4.56 to −1.95; p<0.001) to severe pain (−5.42; 95% CI −7.31 to −3.54; p<0.001)), less education (−2.90; 95% CI −4.17 to −1.64; p<0.001), fewer family income (−1.82; 95% CI −2.77 to −0.87; p<0.001), treated at home (−1.90; 95% CI −3.35 to −0.45; p=0.011) and with lower GSES scores (0.25; 95% CI 0.16 to 0.34; p<0.001).

Conclusions

Patients with limb fractures require rehabilitation but often struggle with adherence to interventions. Strategies that focus on targeted education programmes, pain management and self-efficacy improvement may be considered to guide future research and programmatic approaches to enhance adherence to rehabilitation training.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study provides novel empirical evidence on rehabilitation adherence among Chinese adults with traumatic fractures, a population that has been understudied in existing literature.

• The multiple linear regression model was used to analyse the influencing factors.

• More future studies are required to determine the causal relationship between adherence to postoperative rehabilitation training and the factors influencing it.

Introduction

Bone fractures are a global public health issue and pose a serious social burden,^{1 2} especially with a high incidence of limb fractures.³ Limb fracture is a medical condition that disrupts the continuity of bone, either as a complete or incomplete break commonly caused by falls, automobile accidents or sports injuries.⁴ It can occur and cause various adverse consequences across all age groups, including disability, impaired quality of life, health loss, decreased productivity and high healthcare costs which are major burdens to individuals, families and societies.5,7 According to a systematic analysis of the GBD (Global Burden of Disease) 2019 reported, the Fracture Collaborators noted that the global fracture burden was heavy in 2019, with 178 million (95% uncertainty interval 162 to 196) new fractures and 25.8 million (17.8 to 35.8) YLDs (years of disability).⁸ Unusually, findings from the GBD 2019 Fracture Collaborators showed that the number of YLDs associated with patellar, tibial, fibular or ankle fractures in 2019 was approximately five times higher than the number of YLDs associated with hip fractures.⁸

Rehabilitation relieves pain, increases muscle strength and effectively improves joint range of motion in patients with bone fractures.^{9 10} Additionally, rehabilitation plays a crucial role in reducing hospitalisation and medical expenses.¹¹ In terms of population, China has the largest demand for rehabilitation (460 million), and a large portion of the overall rehabilitation needs are attributed to trauma.¹² A recent study estimated that nearly one billion people have had their lives affected by fractures, amputations, spinal cord injuries, traumatic brain injuries or other impairments. Among them, fractures have the greatest rehabilitation demand, affecting 436 million people.¹³

Rehabilitation therapy adherence is the major predictor of ensuring recovery efficiency and good long-term prognosis, directly affecting the clinical outcome and quality of life of patients with limb fractures.^{14 15} However, non-adherence is common, with 50%–70% of patients either non-adherent or only partially adherent to their rehabilitation programmes, which diminishes the effectiveness of treatment and puts the patient at a higher risk for poor prognosis.^{15 16} Moreover, the outbreak and quarantine of COVID-19 caused healthcare service upheaval, limiting regular and timely rehabilitation treatment in medical facilities. Several different categories of factors influence rehabilitation adherence among orthopaedic patients, including demographics such as age and gender.^{17 18} Relevant disease factors include disease duration and hypoglycaemia, while individual factors include pain, self-efficacy and fear of reinjury.^{19 20} Associated social factors include the availability of professional support and exercise environment.^{21 22} Several systematic reviews have assessed measures to improve patient adherence with rehabilitation training for musculoskeletal disorders.23,25 However, there are comparably few studies on the current situation of adherence and its influencing factors during rehabilitation training in patients with traumatic limb fractures after surgery in China.

In the present study, we aimed to explore postoperative rehabilitation training adherence in adults with traumatic fractures in the Chinese population and to examine cross-sectional correlations between adherence and individual and social-related factors.

Methods

Study design and participants

We conducted a cross-sectional study of postoperative patients with traumatic limb fractures who were followed up from June 2022 to December 2023 at the No. 960th Hospital of PLA in China. The study was reported according to the Strengthening Reporting of Observational Studies in Epidemiology guidelines. All participants signed an informed consent form and that the rights of participants were protected. Inclusion criteria were as follows: (1) age≥18, (2) diagnosed with traumatic limb fractures requiring surgical intervention, confirmed by imaging and meeting the diagnostic criteria for traumatic fractures in the Chinese Guidelines for the Diagnosis and Treatment of Fractures,²⁶ (3) within 1 year of fracture surgery, (4) patients could communicate normally and had no cognitive impairment and (5) were willing to participate in the study. Exclusion criteria were as follows: (1) combined with other fractures, (2) memory impairment that interfered with the performance of daily activities, (3) presence of uncontrolled serious illnesses and (4) pathological fractures (eg, due to osteoporosis or malignancy).

Sample size calculation

Based on the empirical method, the sample size should be more than five times the number of independent variants. The number of independent parameters in this study is 17. Consequently, according to the empirical method, the sample size is considered to be 85.

According to the software method, the sample size was calculated using PASS 2021 software. A sample size of 140 achieves 80% power to detect an effect size (f²) of 0.15 attributable to 6 variables using an F-Test with a significance level (alpha) of 0.05. The calculations assume an unconditional (random X’s) model. Taking into account a 20% sample attrition rate, a total of 168 study participants were needed. To make the results of sample size calculation more convincing, the software method was used in this study.

Measures

The present study used four questionnaires and scales. Patient information was collected using a self-designed general information form based on medical records including (1) social demographic data: patient’s name, age, gender, marital status, occupation, underlying disease, place of residence, whether living alone, education level, income, (2) disease-related data: fracture site, complications, pain, rehabilitation site and (3) daily living habit-related information: exercise habits.

The Exercise Adherence Rating Scale (EARS) is a self-report instrument used to assess adherence to rehabilitation exercise²⁷; the characteristics of parts A and C are optional. In this study, sections B and C related to adherence behaviours and reasons for adherence to exercise were applied separately. Section B consisted of six items and section C consisted of 10 items scored using an ordinal response scale (0=strongly agree, 4=strongly disagree). Total scores ranged from 0 to 64, with higher scores indicating greater adherence. The overall Cronbach’s α coefficient for the scale was 0.81, and the retest reliability ICC=0.97 (0.94–0.98).

Self-efficacy was assessed using the General Self-Efficacy Scale (GSES),²⁸ which consists of 10 items, with each item scoring 1–4 points, and the total score being 10–40 points, with higher scores indicating higher self-efficacy. The Cronbach’s α coefficient for the total scale was 0.982 and the retest reliability was 0.971.

The patient’s ability to perform daily life (ADL)²⁹ was used as an evaluation index for the patient’s functional recovery. ADL was evaluated by the Barthel Index, and a total score of 0–100, the higher the score, the better the patient’s ADL. The internal consistency Cronbach’s alpha coefficient for the simplified version of the Body Mass Index (BMI) was 0.929.

The Visual Analogue Scale (VAS)³⁰ was used to evaluate the intensity of pain on movement, with a total scale score of 0–10, where 0 was no pain, 1–3 was mild pain, 4–6 was moderate pain and 7–10 was severe pain.

Data collection

Face-to-face or telephone interviews were conducted based on patient medical records, general information questionnaires and validated scales. Two postgraduate students were trained before data collection to ensure that they understood the following factors: the purpose and significance of the survey, the structure and content of the questionnaire and how to complete the interviews in a standardised manner. One investigator read through the questions one by one for older adults with poor vision or those who were reluctant to complete them. In this case, subjects answered the questions verbally and the investigator recorded the answers. All questionnaires were field-checked to ensure the accuracy and authenticity of the results.

Statistical analyses

Data processing and analysis were performed using R V.4.3.0 and IBM SPSS V.27.0 (SPSS), and α two-sided p<0.05 was considered statistically significant. Histograms were created to determine the normality of the data. The sociodemographic factors of the participants were described based on the mean and SD of continuous variables, frequency (n) and proportion (%) of categorical variables as well. The effects of sociodemographic and disease-related factors on adherence were compared using χ² tests. Input the factors with p<0.05 in the univariate analysis and include them in the multiple linear regression model. Relationships between adherence to rehabilitation training (independent variable) and demographic data, disease-related factors and exercise habits (dependent variable) were assessed using a multiple linear regression model that adjusted for possible confounders. The intensity of association between the variables was determined by calculating Pearson’s correlation coefficient (r) and the coefficient of determination of correlation (R²). Correlations were determined according to the categorisation proposed by Zou et al³¹: 0=no correlation, 0≥0.20= weak correlation, 0.20≥0.50=moderate correlation, 0.50≥0.80=strong correlation and 0.80≥1.00=perfect correlation.

Patient and public involvement

None.

Results

Participant clinical characteristics

Between December 2022 and December 2023, 182 subjects who had attended the surgical emergency department or were admitted to our hospital because of a trauma limb fracture were invited to participate in the study. Nine were excluded and three refused to participate. Therefore, 170 participants were included in the final analysis (figure 1). Those who agreed to participate received the bone fracture and trauma-related factor assessment. The age of the participants ranged from 18 to 93 years, with a mean age of 41.5 years (SD 15.6). There were 113 (66.5%) participants who were men, and 57 (33.5%) were women. Of these, upper limb fractures were reported by 32 (18.8%) participants, and the majority were lower limb fractures 138 (81.2%). The mean exercise adherence score was 30.31 (SD 5.8). The participants’ clinical characteristics are shown in table 1.

Figure 1. The flow diagram.

Table 1. Demographic characteristics.

Variables	Total sample (n=170)	Variables	Total sample (n=170)
Age (year)		Fracture site
<50	113 (66.5%)	Upper limbs	32 (18.8%)
≥50	57 (33.5%)	Lower limbs	138 (81.2%)
Sex		Complications
Male	113 (66.5%)	None	138 (81.2%)
Female	57 (33.5%)	Yes	32 (18.8%)
Nations		Exercise habits
Han nationality	168 (98.8%)	Yes	44 (25.9%)
Minority nationality	2 (1.2%)	No	126 (74.1%)
Marital status		The intensity of pain on the movement (VAS 1–10 cm)
Married	128 (75.3%)	1–3	11 (6.5%)
Single/divorced	42 (24.7%)	4–6	84 (49.4%)
Occupational activity		7–10	75 (44.1%)
Employed	117 (68.8%)	Rehabilitation setting
Unemployed	53 (31.2%)	Hospitals/rehabilitation centres	93 (54.7%)
Underlying disease		Home	77 (45.3%)
0	130 (76.5%)	Income
1	27 (15.9%)	≥5000	65 (38.2%)
≥2	13 (7.6%)	<5000	105 (61.8%)
Habitation		Education
Urban	75 (44.1%)	Primary or none	70 (41.2%)
Rural	95 (55.9%)	Secondary	47 (27.6%)
Medical insurance		College and above	53 (31.2%)
Yes	89 (52.4%)
None	81 (47.6%)	ADL	84.3±11.3
Living arrangement		GSES	28.5±5.1
Living with family	132 (77.6%)	EARS	30.3±5.8
Living alone or within an organised institution	38 (22.4%)	EARS-RA	18.09±4.18

^*Values are expressed as means (SD) and frequencies.

ADL, activity of daily living; EARS-RA, Exercise Adherence Rating Scale-Reason for Adherence; GSES, General Self-Efficacy Scale; VAS, Visual Analogue Scale.

Correlations

There was a moderately positive correlation between ADL (ρ=0.63, p<0.001), GSES (ρ=0.78, p<0.001) and EARS. Additionally, there was a significant, moderately negative correlation between age (ρ=−0.49, p<0.001), occupational activity (ρ=−0.43, p<0.001), medical insurance (ρ=−0.37, p<0.001), complications (ρ=−0.34, p<0.001), exercise habits (ρ=−0.54, p<0.001), pain (ρ=−0.86, p<0.001), rehabilitation setting (ρ=−0.82, p<0.001), income (ρ=−0.77, p<0.001), education (ρ=−0.83, p<0.001) and EARS (table 2). Although not statistically significant, a weak correlation between marital status (ρ=0.25, p=0.001), sex (ρ=−0.23, p=0.002), underlying disease (ρ=−0.29, p<0.001), habitation (ρ=−0.27, p<0.001) and EARS was found. Thus, to ensure that we accounted for any effect of adherence on rehabilitation training, it was decided to include marital status, sex, underlying disease and habitation into the model.

Table 2. Bivariate correlations between sociodemographic and disease-related factors and rehabilitation training adherence.

Variables	EARS
	ρ	P value
Age	−0.49	<0.001
Sex	−0.23	0.002
Nations	−0.05	0.49
Marital status	0.25	0.001
Occupational activity	−0.43	<0.001
Underlying disease	−0.29	<0.001
Habitation	−0.27	<0.001
Medical insurance	−0.37	<0.001
Living arrangement	0.04	0.624
Fracture site	−0.10	0.218
Complications	−0.34	<0.001
Exercise habits	−0.54	<0.001
Pain (during the movement)	−0.86	<0.001
Rehabilitation setting	−0.82	<0.001
Income	−0.77	<0.001
Education	−0.83	<0.001
ADL	0.63	<0.001
GSES	0.78	<0.001

ADL, activity of daily living; EARS, Exercise Adherence Rating Scale; GSES, General Self-Efficacy Scale.

Univariate linear regression analysis

In the univariate linear regression analysis (table 3), age, sex, marital status, occupational activity, habitation, medical insurance, exercise habits, rehabilitation setting, income and education were all associated with rehabilitation training adherence (p<0.05). Participants who had high scores with ADL (0.32; 95% CI 0.26 to 0.38; p<0.001) and GSES (0.88; 95% CI 0.78 to 0.99; p<0.001) also had good adherence. Concerns about moderate (−6.42; 95% CI −8.24 to −4.60; p<0.001) to severe (−15.20; 95% CI −17.03 to −13.36; p<0.001) pain during movement, with 1 (−3.28; 95%CI −5.58 to −0.98; p=0.006) or ≥2 (−4.98; 95% CI −8.14 to −1.82; p=0.002) underlying disease, and complications (−5.03; 95% CI −7.12 to −2.95; p<0.001) were associated with poor adherence in postoperative rehabilitation training.

Table 3. Univariate linear regression analysis of factors influencing rehabilitation training adherence.

Variables	b	SE	t	P value*	β (95%CI)
Age (year)
<50					−5.95 (−7.55 to −4.35)
≥50	−5.95	0.82	−7.27	<0.001
Sex
Male					−2.81 (−4.60 to −1.02)
Female	−2.81	0.91	−3.08	0.002
Nations
Han nationality					−2.84 (−10.88 to 5.20)
Minority nationality	−2.84	4.10	−0.69	0.490
Marital status
Married					3.26 (1.31 to 5.21)
Single /divorced	3.26	1.00	3.28	0.001
Occupational activity
Employed					−5.27 (−6.97 to −3.57)
Unemployed	−5.27	0.87	−6.09	<0.001
Underlying disease
0
1	−3.28	1.17	−2.80	0.006	−3.28 (−5.58 to −0.98)
≥2	−4.98	1.61	−3.09	0.002	−4.98 (−8.14 to −1.82)
Habitation
Urban					−3.10 (−4.79 to −1.42)
Rural	−3.10	0.86	−3.61	<0.001
Medical insurance
Yes					−4.22 (−5.83 to −2.60)
None	−4.22	0.82	−5.11	<0.001
Fracture site
Upper limbs					−1.39 (−3.60 to 0.82)
Lower limbs	−1.39	1.13	−1.24	0.218
Complications
None					−5.03 (−7.12 to −2.95)
Yes	−5.03	1.06	−4.73	<0.001
Exercise habits
Yes					−7.01 (−8.68 to −5.33)
No	−7.01	0.85	−8.20	<0.001
The intensity of pain on the movement (VAS 1–10 cm)
1–3
4–6	−6.42	0.93	−6.91	<0.001	−6.42 (−8.24 to −4.60)
7–10	−15.20	0.94	−16.23	<0.001	−15.20 (−17.03 to −13.36)
Rehabilitation setting
Hospitals/rehabilitation centres					−9.49 (−10.48 to −8.49)
Home	−9.49	0.51	−18.74	<0.001
Income
≥5000					−9.07 (−10.21 to −7.93)
<5000	−9.07	0.58	−15.53	<0.001
Education
Primary or none
Secondary	−4.97	0.60	−8.28	<0.001	−4.97 (−6.15 to −3.80)
College and above	−11.38	0.58	−19.61	<0.001	−11.38 (−12.52 to −10.24)
ADL	0.32	0.03	10.56	<0.001	0.32 (0.26 to 0.38)
GSES	0.88	0.05	16.19	<0.001	0.88 (0.78 to 0.99)

Bold indicates P<0.05

^*Significance set at p<0.05.

ADL, activity of daily living; GSES, General Self-Efficacy Scale; VAS, Visual Analogue Scale.

Hence, age, sex, marital status, occupational activity, underlying disease, habitation, medical insurance, complications, exercise habits, pain, rehabilitation setting, income, education, ADL and GSES were entered into the final model to determine the influence factors of EARS.

Multivariable linear regression analysis

The ultimate regression model derived from the stepwise multiple linear regression analysis showed that seven variables affected postoperative rehabilitation training adherence among the patients with traumatic fractures: age, marital status, the intensity of pain on the movement, rehabilitation setting, income, education and GSES (table 4). Adherence to postoperative rehabilitation training tended to be lower in participants of older age (−1.30; 95% CI −2.14 to −0.46; p=0.003), and those who had higher pain scores on the movement (moderate pain (−3.25; 95% CI −4.56 to −1.95; p<0.001) to severe pain (−5.42; 95% CI −7.31 to −3.54; p<0.001)), less education (−2.90; 95% CI −4.17 to −1.64; p<0.001), fewer family income (−1.82; 95% CI −2.77 to −0.87; p<0.001), treated at home (−1.90; 95% CI −3.35 to −0.45; p=0.011) with lower GSES scores (0.25; 95% CI 0.16 to 0.34; p<0.001) (table 4). Moreover, single or divorced patients could be worse than the married and were also associated with adherence, but this reached only borderline statistical significance (−0.82; 95% CI −1.58 to −0.005; p=0.037).

Table 4. Multivariate linear regression analysis of factors influencing rehabilitation training adherence.

β (95%CI)	P value*	t	SE	b	Variables
Intercept	30.86	2.36	13.08	<0.001	30.86 (26.24 to 35.49)
Age (year)
<50					−1.30 (−2.14 to −0.46)
≥50	−1.30	0.43	−3.04	0.003
Marital status
Married					−0.82 (−1.58 to −0.05)
Single/divorced	−0.82	0.39	−2.10	0.037
The intensity of pain on the movement (VAS 1–10 cm)
1–3
4–6	−3.25	0.66	−4.90	<0.001	−3.25 (−4.56 to −1.95)
7–10	−5.42	0.96	−5.64	<0.001	−5.42 (−7.31 to −3.54)
Rehabilitation setting
Hospitals/rehabilitation centres					−1.90 (−3.35 to −0.45)
Home	−1.90	0.74	−2.57	0.011
Income
≥5000					−1.82 (−2.77 to −0.87)
<5000	−1.82	0.48	−3.76	<0.001
Education
College and above
Secondary	−0.44	0.48	−0.92	0.359	−0.44 (−1.37 to 0.49)
Primary or none	−2.90	0.65	−4.50	<0.001	−2.90 (−4.17 to −1.64)
GSES	0.25	0.04	5.65	<0.001	0.25 (0.16 to 0.34)

^*Significance set at P<0.05.

GSES, General Self-Efficacy Scale; VAS, Visual Analogue Scale.

The final model significantly accounted for 90.9% of the variance of the exercise adherence and compliance score (F=83.436, p<0.001, adjusted R²=0.898).

Discussion

Low level of postoperative rehabilitation training adherence in patients with traumatic fracture

Nowadays, people’s lifespans are increasing, but often coexist with disability. We believe that rehabilitation services urgently need the attention of medical practitioners to empower people to live healthy lives to the fullest extent possible. According to the latest report from the WHO, rehabilitation is a necessary health resource that should be delivered to the entire population, and it is an indispensable part of attaining universal health coverage.³² Rehabilitation training plays a significant role in the functional recovery of orthopaedic trauma patients, which can effectively accelerate the recovery process of patients, improve joint function and significantly reduce the disability rate. Adherence and compliance to the prescribed rehabilitation regimen are considered vital for achieving successful recovery from limb fractures. However, non-adherence is common, which diminishes the effectiveness of treatment and puts patients at higher risk for poor prognosis.⁹

The findings from this cross-sectional study revealed that participant age, marital status, pain, rehabilitation setting, income, education and GSES were significantly associated with postoperative rehabilitation training adherence. The mean EARS score was 30.31 (SD 5.80) out of 64, and the average EARS-RA score was 18.09 (SD 4.18), which was lower than the study of de Santana et al³³ on rehabilitation adherence in knee osteoarthritis. The underlying reason may be that compared with knee osteoarthritis, patients with limb fractures are more severely injured and have more pronounced pain and, therefore, have lower adherence. However, in contrast with our initial hypothesis, fracture site and exercise habits were not associated with adherence when adjusted for the number of factors and other confounding variables.

Strengthening rehabilitation training is one of the most cost-effective measures to combat the effects of disability. Although this was simply an observational study, these findings, particularly those related to perceived health status and frequency of activity-limiting pain, offer some information about the perceived and real barriers that may need to be addressed to improve adherence.

Influencing factors of postoperative rehabilitation training adherence in patients with traumatic fracture

Sociodemographic factors

According to WHO,³⁴ adherence is modified by factors related to socioeconomic status, healthcare systems, health conditions, therapy and patient-related factors. At the individual level, patients who adhered to rehabilitation training reported younger age (<50 years old), were married, had higher income, higher education level and GSES score with less pain on movement, as well as being treated in hospital or rehabilitation centre than non-adherers.

In our study, the age range is wide, spanning from 18 to 93 years. The reasons for adherence or non-adherence to rehabilitation training may differ considerably between individuals in the lower and higher age groups. Previous studies have shown that compliance with rehabilitation exercises is significantly higher in young and middle-aged patients than in minors and the elderly, probably due to the higher need for functional recovery in this population and the lack of awareness of rehabilitation exercises in minor patients.35,37 Lower adherence to rehabilitation exercises among older adults may be related to their failure to recognise the importance of learning about rehabilitation exercises, lower need for functional recovery, having comorbidities and being too frail to perform rehabilitation exercises.^{38 39} Married patients had higher adherence than single or divorced, possibly because the companionship and support of the partner made the patient more adherent to the exercise. The influence of educational level on the adherence to rehabilitation exercises of patients with fractures is mainly related to the ability to understand knowledge. Evidence suggested that adherence to rehabilitation exercises in fractured patients was positively correlated with their educational level and that patients with a high level of education are more likely to be fully aware of its importance.⁴⁰ Economic level is also an influential factor in exercise adherence in patients after fracture surgery. It has been found that patients with high economic income have better adherence to rehabilitation exercises.⁴¹ The potential exists for individuals with substantial financial means to outsource their daily tasks, such as household chores, transportation services and childcare responsibilities, as well as to more readily secure time off from work. On the contrary, without money, it may be related to the anxiety of low-income patients to seek medical treatment, and anxiety may cause patients to appear resistant. Therefore, the sociodemographic factors of patients such as age, marital status, education level and income should be taken into account when conducting rehabilitation training schedules, so that programmes can be targeted.

The intensity of pain on the movement

Systematic reviews have identified barriers and predictors of adherence to prescribed exercise in patients with chronic musculoskeletal pain.^{15 24 25}

In our cross-sectional study, 49.4% reported moderate pain (VAS score 4–6) and 44.1% severe pain (VAS score 7–10) during exercise. There was a significant association between pain and adherence (p<0.001). Pain is a common symptom in patients with fractures, and it has been found that 80% of patients refuse rehabilitation exercises because of wound pain,⁴² and the adherence of those with severe pain is lower than that of those with less pain.⁴⁰ Rehabilitation exercises inevitably cause or aggravate pain by pulling on the wound, which may make the patient fearful of treatment. In addition, pain also increases the patient’s anxiety, depression and fear, hindering the recovery process.⁴³ However, studies have shown that active rehabilitation not only promotes fracture healing and improves patient prognosis, but also shortens the duration of pain.⁴⁴ Suggesting that rehabilitation exercises cause or aggravate pain is only temporary, and in the long run, early rehabilitation exercises can help relieve pain. Our previous qualitative interviews found that early rehabilitation training after orthopaedic surgery is very important for patients to return to normal life. However, pain during rehabilitation training is rarely a concern in China.⁴⁵

Therefore, the evaluation, monitoring and management of pain in patients should be improved among medical professionals and patients. Medical workers should be aware of the importance of pain management while treating the disability. It is recommended to provide educational programmes related to pain management and strengthen the connection between multiple disciplines to improve how medical workers evaluate and control pain.

Rehabilitation setting

The findings reveal that patients who received rehabilitation training in hospitals or rehabilitation centres were significantly more adherent than those who exercised themselves at home (p=0.011). Consistent with the results of related studies, patients have previously described the coordinating, motivating and guiding role of the rehabilitation therapist in providing high-value evidence-based care to facilitate their return to their rehabilitation goals.⁴⁶ One possible explanation is that within the clinical rehabilitation environment, the characteristics of regular rehabilitation sessions and professional supervision might potentially strengthen rehabilitation adherence. It is noteworthy that the effect of a higher frequency of supervised rehabilitation, particularly in the later phases of rehabilitation, is a particular area of note. 6 months after surgery is the best time to address trauma injuries before returning to normal life,²¹ as the fracture is in a critical stage of healing during this period.⁴⁷ Timely rehabilitation treatment can effectively promote fracture healing and functional recovery. If missed, it may lead to serious problems such as joint adhesion and stiffness, increasing the difficulty of subsequent treatment. Therefore, time is crucial for the recovery of fracture patients.^{41 48}

The prevalence of COVID-19 from 2019 to 2023 may be one of the reasons for the low level of exercise adherence in patients. Residents were required to avoid gathering, and food and daily necessities were delivered to the home by volunteers, leading to a decrease in seeking medical care during the epidemic, so they may not have received timely professional support from medical staff for rehabilitation training.⁴⁹ During this period, some government measures to combat COVID-19 were in effect, which may have influenced the participation of individuals.

At present, the majority of orthopaedic patients in China are unable to complete the full rehabilitation in medical institutions,⁵⁰ and the patients return to the community or family from the hospital, which requires the active development of community and family rehabilitation training. Patient education comprises educational activities designed to influence patient knowledge and health behaviours, enable patients to manage their disease and optimise health outcomes.^{51 52} Therefore, online remote patient education and rehabilitation training should be carried out to improve adherence to rehabilitation treatment, and timely assessment of rehabilitation effects should be carried out to promote rehabilitation training in the community and family. The close integration of clinical treatment and rehabilitation training will enable patients to recover their functions better, faster and more comprehensively and return to society as soon as possible.

General self-efficacy scale

Self-efficacy is an individual’s self-confidence in accomplishing an activity.⁵³ This belief significantly impacts behaviour, activity selection and overall achievement levels. Consequently, patients who doubt their ability to complete rehabilitation training may avoid attempting the prescribed exercises altogether. Self-efficacy also determines the effort individuals invest in a task.⁵³ Self‐efficacy has been studied extensively within cardiac rehabilitation, but less so for patients undergoing musculoskeletal rehabilitation. Initial evidence shows that interventions targeting self‐efficacy are successful in individuals with low back⁵⁴ and knee pain.⁵⁵ Our study found that self-efficacy scores for patients in postoperative fracture rehabilitation were 28.5 (SD 5.1), and rehabilitation exercise adherence was significantly and positively associated with self-efficacy (p<0.001). It can be explained by the fact that higher self-efficacy can positively face and solve the issues, have enough confidence in the prognosis, have a higher degree of acceptance of the rehabilitation exercise, and can be more active in the rehabilitation exercise. In contrast, low self-efficacy is easy to avoid and fear the problem, which reduces the patient’s adherence to the rehabilitation training.⁵⁶

Studies indicate that clinicians who evaluate a patient’s self-efficacy before recommending an exercise programme can tailor these programmes more effectively, for instance, through personalised goal setting and enhanced social support, thereby fostering a patient’s perceived capability.⁵³ A multicentre study in China showed that only 52.26% of orthopaedic surgery patients were aware of postoperative rehabilitation.⁵⁷ Moreover, a higher level of adherence to rehabilitation recommendations in our study was observed in people living with a spouse or a partner, and in patients with a higher education level and income. This may be due to the support that patients receive from a close relationship; they may feel more confident about the treatment than others. Some studies have reported that a lack of a caregiver or a caregiver’s large workload may influence treatment adherence.^{58 59}

Studies indicate that healthcare professionals who assess patients’ self-efficacy before prescribing exercise regimens can more effectively customise these interventions, for instance, through personalised goal setting and enhanced social support, thereby fostering a patient’s perceived capability.⁵³ Therefore, improving patients’ adherence to rehabilitation exercises should improve their self-efficacy. Efforts should be directed towards strengthening patients’ belief in the importance of rehabilitation, thereby increasing their awareness and confidence in completing treatment.

Limitations

This study has several limitations. First, the respondents to the survey were primarily from China, which limits the applicability of the results to other countries due to differences in healthcare services. The survey was completed by adults only, so findings cannot be generalised to the paediatric or adolescent populations. Second, the study was cross-sectional and observational, allowing for the identification of associations between factors influencing adherence and persistence but cannot determine causality. The data collected were largely retrospective, and this introduced the possibility of bias due to recall or rationalisation of patient experiences. Future research needs to consider multiple time points to mitigate this bias. Finally, although we reported characteristics of rehabilitation adherence, we did not include any questions on the content of the rehabilitation quality and level of the participants. Future studies may add relevant factors to validate the extent of impairment for low rehabilitation training adherence.

Despite the limitations of such sample and design, these data do provide scope for further study into the factors related to exercise adherence in patients undergoing postoperative rehabilitation training in orthopaedics.

Conclusions

The results of this survey provide insight into the adult patients’ perspective on the key barriers to and facilitators of traumatic fracture rehabilitation adherence. The most important findings of this survey are that severe pain during movement, rehabilitation settings, GSES scores and the patient’s education and income levels were the strongest factors reported by patients in China that influence their ability to complete rehabilitation as recommended.

Strategies that focus on targeted education programmes and rehabilitation training schedules, pain management and self-efficacy improvement may be considered to guide future research and programmatic approaches to enhance adherence to rehabilitation training for postoperative orthopaedic patients.

The funder had no role in the study design, data collection, analysis, interpretation or report writing.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.