Patient compliance, or adherence, has been widely studied in the medical literature. Most of this research has focused on the adherence of patients to taking prescribed medications. The results of these studies have shown that medication noncompliance is a large problem with far-reaching and costly consequences1.
Within the orthopaedic surgery literature, the majority of research on patient compliance has dealt with bracing. Very few studies have examined patient compliance with postoperative recommendations. This is likely due to difficulty with accurately measuring many of the postoperative recommendations given to patients. Yet for many procedures, postoperative recommendations and restrictions may be crucial to achieving optimal results. Substantial effort is spent studying and reporting on different procedures and the outcomes that they produce without taking into account the role that patient adherence may play in these outcomes. In this study, the compliance of postoperative orthopaedic patients with a strict lower-extremity non-weight-bearing protocol was examined using objective measurements of force placed on the lower extremity.
Materials and Methods
Prior approval for this study was obtained from the institutional review board. Patients who were seen in the office and would be undergoing a foot or ankle surgical procedure were screened for the study. The inclusion criteria were an age of ≥18 years and a unilateral foot or ankle surgical procedure that would require a period of postoperative non-weight-bearing. Patients were excluded if they were nonambulatory at baseline, had dementia, had a neurologic disorder that affected walking or balance (for example, cerebral palsy), had undergone a lower-extremity amputation at the hindfoot or more proximally, had been diagnosed with an axis-II disorder listed in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV), or were unable to maintain non-weight-bearing status at the first postoperative visit.
Patients who were eligible for the study provided informed consent at the preoperative office visit. To minimize the Hawthorne effect (when people change their behavior because they know that they are being observed or studied), patients were not explicitly told that they were going to be monitored for their compliance with the non-weight-bearing restriction. They were told by means of a verbal script that a device would be placed in their cast to monitor their activity. Patients were given explicit directions by their provider and the cast technician with regard to their non-weight-bearing status. Further, the potential complications associated with bearing weight were reviewed. Participation was strictly voluntary, and no compensation was provided to study participants.
Fifty-one adult patients were enrolled in the study between June 2012 and February 2013 (Table I). Preoperatively and postoperatively, all patients received both written and verbal instructions that they must remain non-weight-bearing following their surgical procedure. These written instructions were provided as part of the standard preoperative and postoperative surgical information packets and were no different from the instructions normally provided to non-study patients.
At their first postoperative visit, all 51 participants were switched from an initial postoperative splint into a short leg cast into which pressure-sensitive film (28 to 85 psi [193 to 586 kPa]) (Fujifilm) was embedded. Film with a sensitivity range of 28 to 85 psi (193 to 586 kPa) was chosen because the normal peak pressure measured under the heel of a walking adult falls within this range2. The film turned red when pressure was applied to the film within the pressure sensitivity range. The color density of the red deepened as the applied pressure increased. A full-time certified cast technician, who works exclusively in the foot and ankle surgery clinic, applied all of the casts. This cast technician was aware of the study and was trained to place the casts in a uniform fashion. The film was placed between the initial stockinette layer and the subsequent cotton undercast padding layers of the cast (Fig. 1). Patients were read the study script again at the time of cast application. The film was retrieved at the time of cast removal and was analyzed using a densitometric scanner (Fujifilm) with software capable of determining pressure magnitude and distribution. From this, a mean pressure applied to the film could be measured and used for data analysis.
All patients were followed for a minimum of 3 months postoperatively, and all complications were recorded. A lack of evidence of osseous bridging defined delayed union at 3 months on a radiograph and nonunion at 6 months on a computed tomographic (CT) scan.
Noncompliance was defined as ≥50% of the surface area of the film being exposed to the maximum pressure (>85 psi [>586 kPa]) (Fig. 2). This value corresponded to repetitive full weight-bearing in a small test group. The test group consisted of 6 volunteers who underwent short leg casting and placement of the pressure-sensitive film in identical fashion to the study cohort. Each member of the test group underwent 4 separate casting trials with new film each time. The first trial consisted of non-weight-bearing. The second trial consisted of standing on both feet (50% weight-bearing on each leg) for 30 seconds. The third trial consisted of standing only on the lower limb with the cast (100% weight-bearing on that lower limb) for 30 seconds. The fourth and final trial consisted of walking in the cast for 2 minutes. None of the test group members exceeded 50% of the surface area exposed to maximum pressure in the first 3 trials. However, all 6 test group members exceeded this aforementioned noncompliance cutoff when walking. This strict threshold for noncompliance was used in an effort to negate any potential measurement error from differences in testing (such as padding or film placement differences).
Patient characteristics were collected and were examined to determine if they had a significant effect on weight-bearing. These characteristics were sex, age, language spoken, body mass index (BMI), length of time in the cast, and treating surgeon. Additionally, the time of year in the United States was divided into warmer months (June to October) and colder months (November to March) and was examined for an effect on compliance. Warmer months were defined as the months in which the local daily mean high temperatures for the month are ≥60°F (≥16°C) (range, 61°F to 82°F [16°C to 28°C]). Colder months were defined as the months in which the local daily mean high temperatures for the month are <60°F (<16°C) (range, 36°F to 52°F [2°C to 11°C]).
The mean pressure recorded on each patient’s pressure-sensitive film was used in the statistical analysis to allow for linear regression modeling. No power analysis was performed because the primary aim of this observational study was to determine the noncompliance rate, which is not amenable to a power analysis. The 95% binomial confidence intervals (95% CIs) were determined for the noncompliance rate. An unpaired t test was used for the statistical analysis of sex and time of year. Analysis of variance (ANOVA) was used for the statistical analysis of the treating surgeon. The alpha was set at 0.05 for the t test and ANOVA. Pearson product-moment correlations were used for age, BMI, and length of time in the cast. For the sample size of the study, a correlation of >0.28 was determined to be significant. Then a stepwise multiple linear regression model was utilized to demonstrate predictors of noncompliance with the help of statistical software (SPSS version 10; SPSS). A discriminatory alpha value of <0.05 for the B-statistic delineated significance. The patient’s sex, age, BMI, and number of days in the cast; the treating surgeon; and the time of year spent in the cast were included as variables in the regression equation. The Fisher exact test was used to determine if there was a significant difference in the rate of complications between the compliant and noncompliant patients.
None of the 51 patients in the study were lost to follow-up. The treating surgeon determined when patients no longer required follow-up from their surgical procedure, but all patients were followed for a minimum of 3 months (mean, 9.1 months [range, 3 to 24 months]). The casts were in place for a mean of 24.3 days (range, 7 to 48 days). Fourteen (27.5%) of the 51 patients exceeded the threshold to be considered noncompliant with the non-weight-bearing restriction. The 95% CI for this noncompliance rate was 15.2% to 39.8%. On initial statistical testing, only the time of year was found to have a significant effect on weight-bearing, with weight-bearing being higher in warmer months (June to October) than in colder months (November to March) (p = 0.04). None of the variables examined in the multiple linear regression model were found to have a significant impact on weight-bearing; however, there was a trend toward significance for the impact that the time of year had on weight-bearing (p = 0.06).
Overall, 17 (33.3%) of 51 patients experienced a complication in the postoperative period, including persistent pain, wound problems, delayed union or nonunion, deep venous thrombosis, or painful implants (Table II). Six (42.9%) of the 14 noncompliant patients compared with 11 (29.7%) of the 37 compliant patients experienced a complication. This was not a significant difference (p = 0.51).
The existing literature on compliance with postoperative weight-bearing restrictions has largely focused on partial weight-bearing protocols. Studies have described the challenges of teaching patients partial weight-bearing as well as the poor compliance of patients with partial weight-bearing3,4. However, a series of studies showed improvement with teaching and compliance through the use of biofeedback devices5-10. The reasons behind poor compliance with partial weight-bearing are not entirely clear. Increasing BMI and being touch-down weight-bearing, compared with partial weight-bearing, were shown in one study to increase noncompliance5. There are likely many and variable factors that contribute to compliance with weight-bearing, including comprehension, education, physical, and psychosocial factors.
Compliance in orthopaedics has also been previously studied with regard to postsurgical follow-up appointments, medications, nonoperative bracing, and postoperative restrictions11-22. Many of these studies were limited by their reliance on patient questionnaires and patient-reported compliance rates. It has been shown, when their reports are compared with objective data, that patients overestimate their level of compliance14,18,21. The role that compliance plays in patient outcomes has been shown to be important. For instance, in the nonoperative treatment of clubfeet and congenital hip dysplasia, parental noncompliance has been shown to increase failure rates and to lead to increased surgical intervention13,22.
The main purpose of this study was to determine the rate of patient compliance with a postoperative non-weight-bearing restriction. The rate of compliance was relatively poor at 72.5% (i.e., a noncompliance rate of 27.5%). As a secondary goal, we examined the association of several variables with the extent of compliance. It was found that patients wearing the casts in the warmer months (39.0%) exposed the pressure-sensitive film, on average, 19.1% more than the patients wearing the casts in the colder months (19.9%) (p = 0.04). However, in the linear regression model that adjusted for the other variables, this difference did not quite reach significance (p = 0.06). Additionally, our study found that a higher percentage of patients in the noncompliant group had a postoperative complication, including delayed union or nonunion. However, these differences were not significant, raising the question of whether or not non-weight-bearing is as important for osseous union as many believe it to be. Because the main purpose of our study was to determine compliance, it was underpowered for these secondary goals. For instance, a post hoc power analysis revealed that almost 400 patients would have needed to be enrolled in the study to achieve 80% power with regard to the delayed union or nonunion rate.
It is possible that our observed rate of noncompliance was lower than the actual rate in the general population. It has been previously shown that telling patients that their compliance will be monitored leads to increased compliance rates19. Although we attempted to minimize the Hawthorne effect by blinding patients to the exact purpose of the study, it is difficult to entirely negate this effect when informed consent is obtained. Patients were not explicitly told that their compliance with non-weight-bearing would be monitored, but they were aware that we were monitoring their cast. Because patients knew that their activity would be monitored and they were instructed to be non-weight-bearing, they may have altered their behavior. Noncompliance may also have been underestimated because of the high threshold set for noncompliance. The threshold was set on the basis of the readings from the pressure-sensitive film when the test group walked on the cast. Even the film readings that resulted from standing solely on the limb immobilized in a cast were below the threshold set to define noncompliance.
This study demonstrates a high rate of noncompliance with non-weight-bearing restrictions. In addition to the role that noncompliance may play on outcomes, there are ethical, economic, and medicolegal ramifications of noncompliance that affect patients, physicians, and society as a whole23. The reasons for patient noncompliance are likely complex and need a great deal more investigation. Psychosocial factors such as education, understanding, motivation, support system, cultural beliefs, and underlying psychiatric disease all may play a role23. Further understanding of the causes of noncompliance may help clinicians to develop better solutions and to more effectively target patients at risk for noncompliance.
Do the results of the current study warrant a change in practice? We believe that they do. The findings shine a light on an underappreciated problem in the field of orthopaedics and provide the surgeon with an opportunity to intervene to improve patient outcomes and safety. In fact, the senior authors have changed their practice in response to these findings and now routinely prescribe a session of preoperative physical therapy for gait training in all patients whom they suspect may have a problem with weight-bearing restrictions. The rationale for the physical therapy is discussed with the patient, which also allows for further discussion regarding the importance of compliance.
In conclusion, in this single-blinded study, the noncompliance rate with a postoperative non-weight-bearing restriction was 27.5% (95% CI, 15.2% to 39.8%). The only factor found to have a significant effect on weight-bearing compliance was the time of year that the cast was worn. No significant difference was found in the complication rates between the compliant and noncompliant patients; however, the study was underpowered for this purpose.
Investigation performed at the Department of Orthopaedic Surgery, Brigham and Women’s Hospital, Boston, Massachusetts
A commentary by Kath Bogie, DPhil, is linked to the online version of this article at jbjs.org.
Disclosure: There was no external source of funding provided for this study. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a patent and/or copyright, planned, pending, or issued, broadly relevant to this work.
1. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med. 2005 ;353(5):487–97.
2. Hennig EM, Rosenbaum D. Pressure distribution patterns under the feet of children in comparison with adults. Foot Ankle. 1991 ;11(5):306–11.
3. Hurkmans HL, Bussmann JB, Selles RW, Benda E, Stam HJ, Verhaar JA. The difference between actual and prescribed weight bearing of total hip patients with a trochanteric osteotomy: long-term vertical force measurements inside and outside the hospital. Arch Phys Med Rehabil. 2007 ;88(2):200–6.
4. Tveit M, Kärrholm J. Low effectiveness of prescribed partial weight bearing. Continuous recording of vertical loads using a new pressure-sensitive insole. J Rehabil Med. 2001 ;33(1):42–6.
5. Hustedt JW, Blizzard DJ, Baumgaertner MR, Leslie MP, Grauer JN. Effect of age on partial weight-bearing training. Orthopedics. 2012 ;35(7):e1061–7.
6. Hustedt JW, Blizzard DJ, Baumgaertner MR, Leslie MP, Grauer JN. Current advances in training orthopaedic patients to comply with partial weight-bearing instructions. Yale J Biol Med. 2012 ;85(1):119–25. Epub 2012 Mar 29.
7. Hustedt JW, Blizzard DJ, Baumgaertner MR, Leslie MP, Grauer JN. Is it possible to train patients to limit weight bearing on a lower extremity? Orthopedics. 2012 ;35(1):e31–7.
8. Ruiz FK, Fu MC, Bohl DD, Hustedt JW, Baumgaertner MR, Leslie MP, Grauer JN. Patient compliance with postoperative lower extremity touch-down weight-bearing orders at a level I academic trauma center. Orthopedics. 2014 ;37(6):e552–6.
9. Yu S, McDonald T, Jesudason C, Stiller K, Sullivan T. Orthopedic inpatients’ ability to accurately reproduce partial weight bearing orders. Orthopedics. 2014 ;37(1):e10–8.
10. Hustedt JW, Blizzard DJ, Baumgaertner MR, Leslie MP, Grauer JN. Lower-extremity weight-bearing compliance is maintained over time after biofeedback training. Orthopedics. 2012 ;35(11):e1644–8.
11. Clohisy JC, Kamath GV, Byrd GD, Steger-May K, Wright RW. Patient compliance with clinical follow-up after total joint arthroplasty. J Bone Joint Surg Am. 2008 ;90(9):1848–54.
12. Bergqvist D, Arcelus JI, Felicissimo P; ETHOS investigators. Post-discharge compliance to venous thromboembolism prophylaxis in high-risk orthopaedic surgery: results from the ETHOS registry. Thromb Haemost. 2012 ;107(2):280–7. Epub 2011 Dec 21.
13. McHale KA, Corbett D. Parental noncompliance with Pavlik harness treatment of infantile hip problems. J Pediatr Orthop. 1989 ;9(6):649–52.
14. Morgenstein A, Davis R, Talwalkar V, Iwinski H Jr, Walker J, Milbrandt TA. A randomized clinical trial comparing reported and measured wear rates in clubfoot bracing using a novel pressure sensor. J Pediatr Orthop. 2015 ;35(2):185–91.
15. Silverio LM, Cheung EV. Patient adherence with postoperative restrictions after rotator cuff repair. J Shoulder Elbow Surg. 2014 ;23(4):508–13. Epub 2014 Feb 24.
16. Cuff DJ, Pupello DR. Prospective evaluation of postoperative compliance and outcomes after rotator cuff repair in patients with and without Workers’ Compensation claims. J Shoulder Elbow Surg. 2012 ;21(12):1728–33. Epub 2012 May 29.
17. Lindeman M, Behm K. Cognitive strategies and self-esteem as predictors of brace-wear noncompliance in patients with idiopathic scoliosis and kyphosis. J Pediatr Orthop. 1999 ;19(4):493–9.
18. Takemitsu M, Bowen JR, Rahman T, Glutting JJ, Scott CB. Compliance monitoring of brace treatment for patients with idiopathic scoliosis. Spine (Phila Pa 1976). 2004 ;29(18):2070–4; discussion 2074.
19. Miller DJ, Franzone JM, Matsumoto H, Gomez JA, Avendaño J, Hyman JE, Roye DP Jr, Vitale MG. Electronic monitoring improves brace-wearing compliance in patients with adolescent idiopathic scoliosis: a randomized clinical trial. Spine (Phila Pa 1976). 2012 ;37(9):717–21.
20. Helfenstein A, Lankes M, Ohlert K, Varoga D, Hahne HJ, Ulrich HW, Hassenpflug J. The objective determination of compliance in treatment of adolescent idiopathic scoliosis with spinal orthoses. Spine (Phila Pa 1976). 2006 ;31(3):339–44.
21. Nicholson GP, Ferguson-Pell MW, Smith K, Edgar M, Morley T. The objective measurement of spinal orthosis use for the treatment of adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2003 ;28(19):2243–50; discussion 2250-1.
22. Senthi S, Crawford H, Maxwell T. Outcomes of non-compliance with abduction bracing in the Ponseti treatment for clubfeet in a New Zealand population. Bone Joint Surg Br Ortho Proc. 2011;93-B(SUPP III):376.
23. Sciberras N, Gregori A, Holt G. The ethical and practical challenges of patient noncompliance in orthopaedic surgery. J Bone Joint Surg Am. 2013 ;95(9):e61.