Introduction

With the continued rise in the aging population, knee osteoarthritis has emerged as one of the most common joint disorder among older adults, as well as a common cause of pain, functional limitation, and reduced quality of life among older adults.1 The reported incidence varies according to the diagnostic method: symptomatic venous thromboembolism after total hip or knee arthroplasty occurs in approximately 0.6–1.5% of patients, whereas studies using routine postoperative ultrasonography have detected DVT in 31.3–43.9% of patients after TKA in Asian cohorts.2–4 For patients with advanced knee osteoarthritis who do not obtain sufficient relief from conservative treatment, total knee arthroplasty can substantially improve pain, physical function, and quality of life.5,6 However, TKA patients face multiple potential complications, among which deep vein thrombosis (DVT) poses a particularly high risk.7,8 If not promptly detected and treated, a dislodged thrombus may travel through the bloodstream to the lungs, leading to pulmonary embolism—an extremely severe complication that may be life-threatening.9 Therefore, strengthening DVT risk management in TKA patients is of paramount importance. Failure Mode and Effects Analysis (FMEA) and Root Cause Analysis (RCA) are widely applied risk management tools in the healthcare sector. FMEA is a proactive risk assessment method that analyzes potential failure modes within a system, evaluating their probability of occurrence, severity, and detectability to determine risk priority levels and implement preventive measures accordingly.10–12 In contrast, RCA is a retrospective analysis method designed to identify the root causes of problems and develop effective corrective measures to prevent recurrence.13

In the context of TKA, DVT prevention extends from admission to discharge and depends on timely risk assessment, appropriate prophylaxis, effective communication among healthcare professionals, and patient adherence to preventive measures. Conventional DVT management generally focuses on risk stratification and the use of basic, mechanical, and pharmacological prophylaxis, but it may not systematically identify where failures occur during implementation or why preventive measures are not completed as planned.14 Comparatively, FMEA can prospectively identify and prioritize potential weaknesses throughout the prevention process, whereas RCA can further examine the underlying personnel-, workflow-, training-, and management-related causes of high-risk failure modes.15 Combining these methods may therefore support a more complete improvement process, linking early risk identification with root-cause diagnosis, targeted corrective measures, and subsequent monitoring of implementation. Although FMEA and RCA have each been applied in healthcare quality and safety management, evidence remains limited on their integration into a single, reproducible workflow for perioperative DVT prevention in TKA and on whether this approach can improve both implementation and clinical outcomes. Therefore, this study evaluated whether an integrated FMEA-RCA strategy could reduce postoperative DVT incidence, shorten hospital stay, and improve early knee function, while also assessing implementation fidelity and the reliability of RPN scoring.

Methods General Information

This study adopted a prospective, non-randomized, quasi-experimental (time-sequenced before-after) design to evaluate DVT risk management in patients undergoing TKA. Consecutive eligible patients admitted to the Department of Orthopedics of a tertiary hospital in Hebei Province were assigned to groups according to admission period: patients hospitalized from July 2023 to January 2024 comprised the control group and received conventional DVT risk management based on standard orthopedic preventive protocols, whereas patients hospitalized from February 2024 to August 2024 comprised the study group and received a structured program incorporating Failure Mode and Effects Analysis (FMEA) combined with Root Cause Analysis (RCA). All procedures were performed at the same institution by the same orthopedic surgical team. In addition, anesthetic management and the institutional protocol for perioperative pharmacologic thromboprophylaxis and postoperative rehabilitation were applied consistently across both cohorts, with no protocol-level changes during the study period. Group allocation was determined according to the order of hospital admission. The inclusion criteria were as follows: (1) confirmed diagnosis of knee osteoarthritis through clinical symptoms and imaging examinations, with reference to the Chinese Guidelines for the Diagnosis and Treatment of Osteoarthritis (2021 Edition); (2) meeting surgical indications; (3) undergoing unilateral TKA for the first time; and (4) having normal auditory and visual functions, allowing for normal communication. The exclusion criteria were: (1) severe dysfunction of vital organs; (2) concurrent diseases affecting lower limb mobility; (3) coagulation dysfunction; and (4) a history of thrombosis before surgery or in the past. Comparisons of general demographic data such as age, sex, weight, and education level between the two groups showed no statistically significant differences (P > 0.05), indicating comparability. All patients included in the study signed informed consent forms, and the study was approved by the hospital’s ethics committe.

Control Group

Patients in the control group received routine DVT risk management according to the department’s standard orthopedic protocol. Physicians assessed venous thromboembolism risk using the Caprini Venous Thrombosis Risk Assessment Scale14 at four prespecified time points: within 24 hours of admission, within 24 hours after surgery, when the patient’s condition changed, and before discharge. Based on the Caprini risk category and clinical assessment, patients received basic, mechanical, and pharmacologic prophylaxis according to the institutional protocol. Basic prophylaxis included ankle-pump exercises initiated immediately after the patient regained consciousness from anesthesia, elevation of the affected limb approximately 20–30 cm above heart level, ensuring a daily fluid intake of no less than 1500–2500 mL in the absence of contraindications, and ambulation with walker assistance within 24–48 hours postoperatively, depending on drainage volume and recovery of muscle strength. Mechanical prophylaxis consisted of an intermittent pneumatic compression device, initiated at within 6 hours after surgery and continued for 3 hours twice daily, unless contraindicated by severe lower-extremity arteriosclerosis obliterans, congestive heart failure, severe lower-extremity edema or dermatitis, or recent skin grafting. Pharmacologic prophylaxis consisted of enoxaparin at 4000 IU subcutaneously once daily, initiated 12 hours after surgery and continued for 14 days. The regimen was withheld or modified in patients with active bleeding or high bleeding risk, severe renal impairment (creatinine clearance <30 mL/min), thrombocytopenia (<50×109/L), or known allergy to heparin or rivaroxaban. The same protocol was applied during both study periods.

Study Group

On the basis of routine nursing care, the study group received a DVT risk-management intervention integrating FMEA with RCA.

Establishing a Management Team: An eight-member management team was formed, consisting of one department head, one head nurse, two physicians, two team leaders, and two nurses. Among them, two held senior professional titles, five had intermediate titles, and one held a junior title. The admission criteria for team members were as follows: ① Bachelor’s degree or higher; ② Extensive work experience, with ≥8 years of service; ③ Intermediate or higher professional title; ④ Clear understanding of the responsibilities related to this research project; ⑤ Familiarity with the DVT prevention process for TKA patients in the department, and successful completion of systematic training and assessment on FMEA and RCA. Before implementation, all team members completed standardized training covering the DVT-prevention pathway, FMEA scoring criteria, RPN calculation, RCA procedures, and documentation requirements, followed by case-based practice and assessment to confirm consistent understanding of the intervention procedures. Developing a DVT Prevention Process for TKA Patients: The management team formulated the prevention process through literature review and discussions based on practical experience. Using the brainstorming method, the DVT prevention process for TKA patients was divided into four main stages: admission, intraoperative, postoperative, and discharge, along with seven sub-processes. Details are provided in Table 1. Failure Mode Analysis: The team members list all possible failure modes for each subprocess. Each member independently rates the severity (S), frequency of occurrence (O), and likelihood of detection (D) for each failure mode. Severity (S) refers to the degree of impact if a failure mode occurs, with a rating scale of 1 to 10. Frequency of occurrence (O) represents the probability of a failure mode occurring, rated from 1 to 10. Likelihood of detection (D) indicates the ability to detect a failure mode or its cause during service, rated from 1 to 10.16 The team aggregates individual scores and takes the average as the final value. The Risk Priority Number (RPN) is then calculated for each failure mode using the formula: RPN = S × O × D A higher RPN indicates a higher risk and a more urgent need for resolution.17,18 Studies have shown19,20 that when the RPN score exceeds 125, the failure mode is considered high-risk and requires immediate improvement measures. See Table 2 for details. Of the 14 potential failure modes identified across the seven subprocesses, eight had a mean RPN greater than 125 and were therefore selected for further RCA and development of targeted improvement measures. These included failure to complete or accurately perform risk assessment at admission and after surgery, and insufficient knowledge or implementation of preoperative and postoperative preventive measures. Formulating Improvement Measures: Through focused discussions, the team ultimately identified eight high-risk failure modes, categorized into four major groups: lack of evaluation (1A+4A), inaccurate evaluation (1B+4B), lack of preventive measures (2A+5A), and inadequate implementation of preventive measures (2B+5B). Using brainstorming and the RCA method, the team explored potential failure factors related to these failure modes from six aspects: personnel, equipment, materials, methods, environment, and measurement.13 Subsequently, the team conducted a fishbone diagram analysis, following these principles:21 ① If this cause did not exist, would the issue still occur? ② If this cause were corrected or eliminated, would a similar issue still arise under the same triggering conditions? If the answer was “no,” the cause was identified as a root cause; if “yes,” it was considered a proximate cause. Based on this analysis, the root cause of each high-risk failure mode was determined (see Table 3), and corresponding improvement measures were developed.

Table 1 DVT Prevention Process for Total Knee Arthroplasty Patients

Table 2 Comparison of Potential Failure Modes and RPN Scores (Unit: Points)

Table 3 Root Causes of High-Risk Failure Modes

The root causes and corresponding core improvement measures for each high-risk failure mode are presented in Supplementary Table S1 (Root causes and core improvement measures for high-risk failure modes [expanded from Table 3]). The training outlines (PBL/CBL), ISBAR handover checklist template, and HAPA-based patient education materials used during implementation are provided in Supplementary Appendix C (Training documentation record [PBL/CBL; audit-level only]), Supplementary Appendix A (ISBAR handover checklist template [for DVT risk management]), and Supplementary Appendix B (Patient education record [HAPA-informed; documentation template]), respectively.

Process Measures: a. Establishing a Comprehensive Supervision System and Dynamic Adjustment Mechanism: A full-process supervision system should be developed, covering the formulation, execution, monitoring, and feedback of preventive measures. The responsibilities of supervision entities at each stage should be clearly defined. The head nurse or quality control team members should conduct regular inspections and evaluations of the implementation of preventive measures, promptly identifying issues and ensuring corrective actions are taken. Additionally, prevention plans should be dynamically adjusted based on changes in patient conditions, treatment progress, and the effectiveness of implemented preventive measures, ensuring continuous effectiveness and adaptability. b. Innovating Health Education Methods and Enhancing Education Quality: Modern information technology should be leveraged to develop diverse health education tools. A departmental WeChat official account and video channel should be established to provide personalized health education content, including video demonstrations of preventive measures, scheduled reminders, and online consultation services, catering to the different needs of patients and their families. Furthermore, training for health educators should be strengthened to improve their knowledge and delivery skills, ensuring that health education content is accurate, easy to understand, and practical, effectively guiding patients and families in correctly implementing preventive measures.

Integrated FMEA-RCA implementation and fidelity assessment: The integrated FMEA-RCA program was implemented as shown in Figure 1. Briefly, a multidisciplinary management team first mapped the department’s DVT prevention workflow across the perioperative pathway (admission, intraoperative, postoperative, and discharge) and identified potential failure modes within each subprocess. Each failure mode was scored for severity (S), occurrence (O), and detectability (D), and the risk priority number (RPN) was calculated (RPN = S × O × D). To ensure scoring consistency, all team members underwent standardized training before formal scoring. Standardized cases were used for practice and discussion until consensus on scoring criteria was achieved. Inter-rater reliability was then evaluated using the intraclass correlation coefficient (ICC); the ICC was 0.87, indicating good agreement among raters. Failure modes with RPN > 125 were prioritized as high risk and subsequently underwent root cause analysis (RCA). RCA was conducted using structured brainstorming and fishbone (Ishikawa) analysis across personnel, equipment, materials, methods, environment, and measurement, and root causes were confirmed through team consensus. Targeted corrective actions were then developed, implemented, and reviewed iteratively, with effectiveness evaluated using both outcome indicators (eg, DVT incidence and length of stay) and predefined process indicators of compliance. In addition, implementation fidelity was monitored using measurable process indicators such as completion of Caprini risk assessments at prespecified time points, documentation of risk-stratified prophylaxis (mechanical and pharmacologic, as applicable), completion of standardized patient education, and supervision/documentation of functional exercise. The detailed definitions and data sources of these compliance indicators are presented in Supplementary Table S2 (Process indicators used to monitor implementation fidelity [compliance]).

Figure 1 Integrated FMEA-RCA workflow for DVT risk management in TKA. The program followed a closed-loop sequence: team formation and workflow mapping; FMEA scoring of failure modes (S, O, D) with RPN calculation (RPN = S × O × (D) and prioritization of high-risk items (RPN > 125); RCA of prioritized failure modes using fishbone analysis; implementation of targeted improvements; and evaluation using RPN values, DVT incidence, length of stay, and HSS knee score (preoperative and 6-week postoperative). Measures were standardized when effective and reanalyzed when not sustained.

Abbreviations: DVT, deep vein thrombosis; TKA, total knee arthroplasty; FMEA, failure mode and effects analysis; RCA, root cause analysis; RPN, risk priority number; S, severity; O, occurrence.

Compliance monitoring. To minimize performance variation across the perioperative period, the department used a standardized DVT-prevention operating manual that specified required actions and documentation criteria at each time point (admission, within 24 h postoperatively, condition change, and discharge). During the study, the head nurse (or designated quality-control nurse) conducted scheduled spot-check audits twice weekly using a structured checklist aligned with the predefined process indicators. Adherence was recorded as the proportion of required items completed per patient per audit cycle based on the predefined process indicators listed in Supplementary Table S2 (Process indicators used to monitor implementation fidelity [compliance]), and deviations were fed back to the clinical team for timely correction.

Evaluation Indicators

① To evaluate changes in process risk, the same eight high-risk failure modes and identical FMEA scoring criteria were applied to both groups. For the control group, the management team retrospectively reviewed the available nursing records, Caprini assessment forms, preventive-measure records, handover documentation, and patient-education records from the routine-care period. Team members independently scored the severity, occurrence and detectability of each failure mode using the same 10-point criteria applied to the intervention group, and the mean values were used to calculate the final RPN. For the intervention group, RPN scoring was completed during implementation of the integrated FMEA-RCA program. ② Comparison of DVT incidence between the two groups. DVT was assessed using clinical evaluation (eg, lower-limb swelling, pain, increased skin temperature, and Homans sign) and was confirmed by color Doppler ultrasonography. The final diagnosis was independently confirmed by at least two senior clinicians blinded to group allocation. ③ Comparison of the hospital stay duration between the two groups. ④ Comparison of knee function recovery between the two groups. The Hospital for Special Surgery (HSS) knee score was assessed preoperatively and at 6 weeks postoperatively, including pain (30 points), function (22 points), range of motion (18 points), muscle strength (10 points), flexion deformity (10 points), and stability (10 points). The total score ranges from 0 to 100 points, with higher scores indicating better knee function.

Statistical Analysis

Statistical analysis was conducted using SPSS 19.0 software. The sample size was estimated based on data from a preliminary pilot study. Using a two-sided significance level of 0.05 and 80% power, and assuming that the intervention would reduce the incidence of postoperative DVT from 20% to 5%, the minimum required sample size was 64 participants per group. To account for potential attrition or incomplete outcome data, we planned to enroll at least 68 participants in each group. Continuous variables were assessed for normality using the Shapiro–Wilk test and for homogeneity of variances using Levene’s test. Normally distributed continuous variables are presented as mean ± standard deviation () and were compared between groups using independent-samples t-tests; within-group pre-post comparisons were performed using paired t-tests. Categorical variables are presented as counts and percentages and were compared using the χ2-test.

For comparisons of RPN values across the eight prespecified high-risk failure modes, multiplicity was controlled using the Bonferroni method, with the adjusted significance threshold set at P < 0.00625 (0.05/8).

To account for potential confounding due to the non-randomized design, we performed multivariable binary logistic regression for postoperative DVT. The model included group assignment (intervention vs control) and clinically relevant covariates, including age, BMI, and baseline Caprini score. The corresponding adjusted odds ratios (aORs) with 95% confidence intervals (CIs) are reported. For continuous outcomes (length of stay and HSS score), between-group mean differences with 95% CIs are reported. A two-sided Pvalue < 0.05 was considered statistically significant unless otherwise specified.

Ethics Statement

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Research Ethics Committee of The Second Hospital of Hebei Medical University (Approval No. 2023-R291). Institutional permission was granted, and written informed consent was obtained from all participants prior to study enrollment.

Results Baseline Characteristics

A total of 136 patients were included (68 in each group). Baseline demographic and clinical characteristics were comparable between groups, including age, sex, body mass index, education level, comorbidities, and baseline Caprini risk score (all P > 0.05; Table 4).

Table 4 Baseline Characteristics of the Two Groups

Comparison of RPN Values for the Eight High-Risk Failure Modes in the DVT Prevention Process Between the Two Groups

RPN values for all eight prespecified high-risk failure modes were lower in the intervention group than in the control group, and all between-group differences remained statistically significant after Bonferroni adjustment (adjusted significance threshold: P < 0.00625). Across the eight failure modes, the mean relative reduction in RPN was 36.29% (range, 16.03% to 59.77%). The largest reduction was observed for poor implementation of postoperative prevention measures (5B), with RPN decreasing from 247.88 ± 37.50 to 99.73 ± 27.27. The detailed results are shown in Table 5.

Table 5 Comparison of RPN Values for Eight High-Risk Failure Modes in the DVT Prevention Process (Points)

Comparison of DVT Incidence Rates Between the Two Groups

The incidence of postoperative DVT was lower in the intervention group than in the control group (5.88% vs 23.53%; χ2 = 8.442, P = 0.004). After adjustment for age, BMI, and baseline Caprini score in a multivariable logistic regression model, the intervention group remained at lower odds of DVT (aOR = 0.24, 95% CI: 0.08–0.72; P = 0.011) (Table 6).

Table 6 Comparison of DVT Occurrence Between Groups (n, %)

Comparison of Hospital Stay Duration Between the Two Groups

The hospital stay duration was shorter in the intervention group than in the control group (7.99 ± 2.06 vs 11.20 ± 4.21 days; t = 5.647, P < 0.001). The between-group mean difference was −3.21 days (intervention minus control), with a 95% CI of −4.33 to −2.09 (Table 7).

Table 7 Comparison of Length of Stay Between Groups (Days)

Comparison of Knee Function Recovery Between the Two Groups

Preoperative HSS scores were comparable between groups (42.50 ± 6.30 vs 41.80 ± 7.10; t = 0.623, P = 0.534). At 6 weeks postoperatively, HSS scores were higher in the intervention group than in the control group (82.30 ± 5.60 vs 76.80 ± 6.90; t = 4.327, P < 0.001). The between-group mean difference at 6 weeks was 5.50 points (intervention minus control), and the corresponding 95% CI is reported in Table 8.

Table 8 Comparison of HSS Knee Scores Between the Two Groups (Points)

Discussion Integration of FMEA and RCA for Identifying Key Weak Links in DVT Prevention

By integrating FMEA with RCA, this study systematically identified four key vulnerable areas in the DVT prevention process for patients undergoing TKA: failure to perform risk assessment, inaccurate risk assessment, insufficient understanding of preventive measures, and inadequate implementation of prevention strategies. Through in-depth analysis of underlying personnel-, process-, and training-related factors, targeted improvement measures were developed and implemented. These interventions were associated with significant reductions in RPN values for high-risk failure modes, a lower incidence of postoperative DVT, a shorter length of hospital stay, and improved early knee function recovery. Collectively, these findings suggest that the integrated risk-management approach contributed to a more comprehensive and individualized nursing care process, with potential benefits for both patient outcomes and overall care efficiency.

Optimization of Risk Assessment Processes to Enhance Prevention Implementation

Several specific mechanisms may explain these improvements. The dual nurse-physician verification system for Caprini risk assessment reduced subjectivity associated with single-person evaluations, improved assessment accuracy, and helped ensure that high-risk patients were identified in a timely manner and received appropriate preventive interventions. Patient education based on the Health Action Process Approach (HAPA) addressed behavior change across three stages, motivation, action planning, and execution monitoring, thereby enhancing patients’ understanding of preventive measures and their willingness to adhere to interventions such as compression stocking use and early mobilization. Improved adherence may, in turn, reduce venous stasis and hypercoagulability, lowering DVT risk from a pathophysiological perspective. Standardized ISBAR handover structured the transfer of information related to patient identity, current status, assessment, and recommendations, reducing information loss, accelerating clinical decision-making, and potentially preventing delays that contribute to prolonged hospitalization. In addition, electronic reminders and dual-check mechanisms helped reduce missed assessments at the process level and supported continuity of preventive care.

Reduction in DVT Incidence and Improvement in Healthcare Quality

The present findings are consistent with previous studies showing that structured FMEA- and RCA-based approaches can identify weaknesses in clinical processes and support targeted quality-improvement measures.22–25 The contribution of this study lies in applying an integrated FMEA-RCA approach to the specific context of perioperative DVT prevention in TKA patients and linking process optimization to clinically relevant outcomes. By refining the risk-assessment workflow and establishing multi-channel reminders and dual-verification mechanisms, assessment accuracy and timeliness were improved, enabling early identification and intervention for high-risk patients. Nevertheless, the marked difference in DVT incidence should be interpreted cautiously because adjustment for age, BMI, and baseline Caprini score cannot exclude residual confounding or other differences between the two admission periods. HAPA-based education combined with strengthened supervision improved adherence to preventive measures, supporting their effective execution. Hospital length of stay is multifactorial and may be influenced by postoperative complications and individual discharge readiness; although our perioperative and rehabilitation pathways were unchanged across cohorts, these factors were not modeled explicitly, and residual confounding cannot be excluded. The observed reduction in length of stay in the intervention group may be related to fewer DVT-related complications as well as improved care coordination associated with standardized handover. Furthermore, the higher 6-week postoperative HSS scores in the intervention group suggest that the integrated FMEA-RCA strategy may also have a positive effect on functional recovery.

While the observed improvements are encouraging, they should be interpreted considering plausible alternative explanations. First, as with many quality-improvement initiatives, heightened attention to DVT prevention during implementation may itself increase adherence independent of the specific components of the FMEA-RCA package (a performance effect). Second, since several process indicators rely on chart documentation, improvements may partly reflect better recording and audit readiness rather than a purely clinical change in practice. Third, although we attempted to reduce between-period variability through consistent institutional pathways and statistical adjustment, time-varying factors not captured in the dataset may still contribute to the observed differences. Accordingly, the findings support an association between the integrated strategy and improved process and clinical outcomes, but they do not establish definitive causality, and external validation in multicenter studies would strengthen generalizability.

Novelty and Contribution

The main contribution of this study is the development of an operational risk-management model for perioperative TKA care that links prospective identification and prioritization of high-risk failure modes with root-cause diagnosis, targeted corrective measures, and measurable monitoring of implementation. Implementation was monitored using predefined compliance indicators, and the reliability of RPN scoring was supported by good agreement among raters. The intervention was associated with lower process risks, a reduced incidence of DVT, a shorter hospital stay, and better early postoperative knee function. This workflow may also be useful for orthopedic nursing teams in other hospitals because it provides a structured process for identifying weaknesses in DVT prevention, analyzing their causes, implementing targeted measures, and monitoring adherence. Although the specific failure modes and improvement measures may differ among institutions, the same workflow can be adapted to local staffing arrangements, clinical procedures, and DVT-prevention practices.

Limitation

This study used a non-randomized, time-sequenced before-and-after design, which is appropriate for clinical quality-improvement research but carries inherent risks of selection bias, temporal bias, and unmeasured confounding. The institutional setting, surgical team, anesthesia management, pharmacologic thromboprophylaxis protocol, and postoperative rehabilitation pathway remained consistent across the two study periods, with no protocol-level changes. Nevertheless, secular changes, including variations in clinical workflow, staffing, and patient case mix, cannot be fully excluded. Although baseline characteristics were compared and multivariable adjustment was performed for key confounders, including age, BMI, and baseline Caprini score, residual confounding may remain. Therefore, the observed differences cannot be attributed solely to the FMEA-RCA intervention, and future multicenter randomized or cluster-randomized studies are required to provide stronger causal evidence.

Conclusion

This quasi-experimental study examined an integrated FMEA-RCA approach for perioperative DVT risk management in patients undergoing TKA. The approach identified key weaknesses in the DVT-prevention process and was associated with lower postoperative DVT incidence, shorter hospital stay, and higher HSS scores at 6 weeks. It also provides a structured workflow linking risk identification, root-cause analysis, corrective measures, and implementation monitoring. Given the single-center, non-randomized design, these findings should be interpreted with caution and require confirmation in multicenter randomized studies.

Abbreviations

TKA, Total Knee Arthroplasty; DVT, Deep Vein Thrombosis; FMEA, Failure Mode and Effects Analysis; RCA, Root Cause Analysis; RPN, Risk Priority Number; ISBAR, Identification, Situation, Background, Assessment, Recommendation; HAPA, Health Action Process Approach; PBL, Problem-Based Learning; CBL, Case-Based Learning; ICU, Intensive Care Unit; SPSS, Statistical Package for the Social Sciences.

Data Sharing Statement

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics Approval and Consent to Participate

This study was approved by the Institutional Ethics Committee. Written informed consent was obtained from all participants prior to inclusion in the study. All procedures were performed in accordance with the ethical standards of the institutional research committee and the Declaration of Helsinki.

Funding

This work was supported by the 2024 Medical Science Research Project of the Health Commission of Hebei Province (Grant No. 20242113).

Disclosure

The authors declare that they have no competing interests.

References

1. Hunter DJ, March L, Chew M. Osteoarthritis in 2020 and beyond: a lancet commission. Lancet. 2020;396(10264):1711–13. doi:10.1016/S0140-6736(20)32230-3

2. Gao J, Xue Z, Huang J, Chen L, Yuan J, Li J. Risk of deep vein thrombosis (DVT) in lower extremity after total knee arthroplasty (TKA) in patients over 60 years old. J Orthopaedic Surg Res. 2023;18(1):865. doi:10.1186/s13018-023-04339-7

3. Itou J, Kuwashima U, Itoh M, Okazaki K. Robotic-assisted total knee arthroplasty is not associated with increased risk of postoperative deep vein thrombosis. J Exp Orthopaedics. 2023;10(1):65. doi:10.1186/s40634-023-00628-6

4. Santana DC, Emara AK, Orr MN, et al. An update on venous thromboembolism rates and prophylaxis in Hip and knee arthroplasty in 2020. Medicina. 2020;56(9):416. doi:10.3390/medicina56090416

5. Ayers DC, Yousef M, Yang W, Zheng H. Age-related differences in pain, function, and quality of life following primary total knee arthroplasty: results from a FORCE-TJR (Function and outcomes research for comparative effectiveness in total joint replacement) cohort. J Arthroplasty. 2023;38(7):S169–S176. doi:10.1016/j.arth.2023.04.005

6. Woodland N, Takla A, Estee MM, et al. Patient-reported outcomes following total knee replacement in patients aged 65 years and over—a systematic review. J Clin Med. 2023;12(4):1613. doi:10.3390/jcm12041613

7. Dai WL, Lin ZM, Shi ZJ, Wang J. Venous thromboembolic events after total knee arthroplasty: which patients are at a high risk? J Knee Surg. 2020;33(10):947–957. doi:10.1055/s-0039-1688962

8. Tong K, Liu H, Qin J, et al. Current status survey of the extramural hospital management of venous thromboembolism after total hip and knee arthroplasty in China. BMC Musculoskelet Disord. 2021;22(1):787. doi:10.1186/s12891-021-04663-1

9. Arnindita JN, Mahyuddin MH, Tri Saputra PB, Putra RM. Identifying pulmonary embolism - high risk patients after total knee replacement: systematic review. J Pak Med Assoc. 2024;74(6):S41–s50. doi:10.47391/JPMA.S6-ACSA-08

10. Pueyo-López C, Sánchez-Cuervo M, Vélez-Díaz-Pallarés M, Ortega-Hernández-Agero T, Salazar-López de Silanes EG. Healthcare failure mode and effect analysis in the chemotherapy preparation process. J Oncol Pharm Pract. 2021;27(7):1588–1595. doi:10.1177/1078155220962189

11. Li G, Xu B, He RX, Zhang SX. Using healthcare failure mode and effect analysis to reduce intravenous chemotherapy errors in chinese hospitalized patients. Cancer Nurs. 2017;40(2):88–93. doi:10.1097/NCC.0000000000000348

12. Lin L, Wang R, Chen T, Deng J, Niu Y, Wang M. Failure mode and effects analysis on the control effect of multi-drug-resistant bacteria in ICU patients. Am J Transl Res. 2021;13(9):10777–10784.

13. Sluggett JK, Lalic S, Hosking SM, et al. Root cause analysis of fall-related hospitalisations among residents of aged care services. Aging Clin Exp Res. 2020;32(10):1947–1957. doi:10.1007/s40520-019-01407-z

14. Krauss ES, Segal A, Dengler N, Cronin M, Pettigrew J, Simonson BG. Utilization of the Caprini score for risk stratification of the arthroplasty patient in the prevention of postoperative venous thrombosis. In: Seminars in Thrombosis and Hemostasis: 2022. Thieme Medical Publishers, Inc.; 2022:407–412.

15. Shaqdan K, Aran S, Besheli LD, Abujudeh H. Root-cause analysis and health failure mode and effect analysis: two leading techniques in health care quality assessment. J Am College Radiol. 2014;11(6):572–579. doi:10.1016/j.jacr.2013.10.024

16. Zhang L, Zeng L, Yan Y, Hang Q. Application of the healthcare failure mode and effects analysis system to reduce the incidence of posture syndrome of thyroid surgery. Medicine (Baltimore). 2019;98(51):e18309. doi:10.1097/MD.0000000000018309

17. Lv S, Sun Y, Zhang J, Jin T, Hu X. Application of the FMEA method in improving the quality management of emergency complete blood count testing. Lab Med. 2023;54(6):574–581. doi:10.1093/labmed/lmad002

18. Chiereghin A, Squillace L, Pizzi L, Bazzani C, Roti L, Mezzetti F. Applying the healthcare failure mode and effects analysis approach to improve the quality of an organised colorectal cancer screening programme. J Med Screen. 2024;31(2):70–77. doi:10.1177/09691413231197300

19. Xin C, Guo X, Ren SY. Optimization and effect evaluation of in-hospital emergency process for extracorporeal cardiopulmonary resuscitation. Chin J Nurs. 2021;56(08):1138–1144.

20. Zhang J, Huang -B-B, Yu T-H, Liu X-Z, Gu J. Optimization of in-hospital resuscitation process for patients with acute ischemic stroke based on health failure mode and effect analysis. J Nurs. 2020;27(4):16–20.

21. Kumah A, Nwogu CN, Issah A-R, et al. Cause-and-effect (fishbone) diagram: a tool for generating and organizing quality improvement ideas. Global J Quality Safety Healthcare. 2024;7(2):85–87. doi:10.36401/JQSH-23-42

22. Dhawan S, Shubham S, Puri N, et al. Improving prescription completeness and reducing prescription errors in a neonatal intensive care unit: a quality improvement initiative from a tertiary care centre in North India. Cureus. 2025;17(9).

23. Shin S, Kang D, Walmsley RS. Post-endoscopy esophageal adenocarcinoma and root cause analysis in Auckland, New Zealand. Endoscopy International Open. 2025;13(continuous publication). doi:10.1055/a-2676-3883

24. Tang X, Yue J, Chen Y, et al. The application of artificial airway care for postoperative weaning patients with cerebral hemorrhage by failure mode and effect analysis mode. Front Med. 2025;12:1645973. doi:10.3389/fmed.2025.1645973

25. Wu H, Xu P, Han B, Xu J. Analysis of nosocomial infection management in the neonatal intensive care unit based on logistic regression combined with FMEA: a retrospective cohort study. J Hosp Infect. 2025;169:85–92. doi:10.1016/j.jhin.2025.11.013