Total hip arthroplasty is claimed to be one of the most highly successful surgical procedures, as it consistently restores function, relieves pain, and improves quality of life for those patients having end-stage degenerative hip disease1-4. Traditionally, strict postoperative outpatient therapy has been administered to patients undergoing total hip arthroplasty in an effort to optimize return of function5-12. Although such a practice may have been justified in the early era of arthroplasty, improvements in surgical and anesthesia techniques in recent years bring into question the value and cost-effectiveness of outpatient therapy after total hip arthroplasty. In fact, prior studies have demonstrated that informal home exercise may be effective in allowing early return of function to patients undergoing total hip arthroplasty5,8,9,13. To our knowledge, to date, no Level-I trial exists comparing an informal home program and outpatient therapy. This is a critical issue that requires examination, as outpatient therapy services after total hip arthroplasty accounted for $180.4 million of costs for U.S. Medicare patients alone in 200914,15. Given the uncertain benefit of utilizing such a costly ancillary service, the efficacy of outpatient therapy in improving outcomes after total hip arthroplasty serves as a relevant and practical clinical question. The purpose of this randomized controlled trial was to compare the effectiveness of an unsupervised home exercise program and outpatient therapy in terms of restoring function for patients undergoing unilateral total hip arthroplasty.
Materials and Methods
This study was a prospective, single-center, 2-arm, parallel-group, randomized controlled trial. The study proposal was approved by our institutional review board. The trial was registered at ClinicalTrials.gov (NCT02687945).
Eligible participants were between 18 and 80 years of age undergoing primary, unilateral total hip arthroplasty for osteoarthritis. The following patients were excluded: those with inflammatory or posttraumatic arthritis, those with a history of septic arthritis of the involved hip, and those undergoing revision total hip arthroplasty or conversion total hip arthroplasty with removal of previously implanted components. Additionally, patients requiring discharge to an acute rehabilitation center, skilled nursing facility, convalescent home, or long-term care facility were excluded. Between September 2013 and October 2015, 120 patients (66 men and 54 women) were enrolled. Sixty patients were allocated to formal outpatient physical therapy and 60 patients were allocated to unsupervised home exercise. Demographic data including age, body mass index (BMI), sex, and Charlson Comorbidity Index were collected preoperatively. The Charlson Comorbidity Index was provided both as an independent comorbidity score and as an age-adjusted score; the age-adjusted Charlson Comorbidity Index was established in 1994 as a combined score that accounts for both age and comorbidity16.
An Excel random number generator (Excel 2013; Microsoft) was used to determine the allocation order using sequentially numbered sealed envelopes that were opened just prior to the surgical intervention, at which time patients were informed of their group allocation. Separate individuals completed the random allocation sequence, patient enrollment, and outcome assessment. As all outcomes were patient-reported, outcome assessors were not blinded to treatment group.
Five adult joint reconstruction surgeons performed the procedures at Thomas Jefferson University Hospital, Philadelphia, Pennsylvania. The surgical approach in all cases was either the direct lateral or the direct anterior, as previously described17,18. Standardized perioperative protocols were used for all patients. Preoperatively, patients received oral acetaminophen (975 mg), pregabalin (75 mg), and celecoxib (400 mg) or ketorolac (30 mg) within 2 hours of the surgical procedure. All patients received uncemented acetabular and femoral components, which were implanted under spinal anesthesia with 0.5% bupivacaine. Surgical drains were not used. Postoperatively, standing doses of oral acetaminophen (650 mg) every 6 hours, pregabalin (75 mg) every 12 hours, and intravenous ketorolac (30 mg) every 6 hours were administered. Oral narcotics (mainly oxycodone 10 mg) and tramadol (50 mg) were administered for residual breakthrough pain and after discharge from the hospital. Aspirin (81 mg or 325 mg twice daily) or warfarin (target international normalized ratio, 1.5 to 2.0) was administered as prophylaxis for prevention of thromboembolism. All patients were mobilized on the day of the surgical procedure. Postoperative hip precautions were not utilized, as is standard at our institution19.
All patients received daily inpatient physical therapy and occupational therapy until the time of hospital discharge. The formal outpatient physical therapy group received 2 weeks of in-home physical therapy followed by formal outpatient therapy, with 2 to 3 weekly sessions for an additional 8 weeks after the surgical procedure. Additionally, patients were provided with a list of suggested physical therapy exercises to be performed at home. The unsupervised home exercise group followed a 10-week unsupervised home exercise program based on a detailed physical therapy manual that was provided to patients prior to discharge. This manual provided images and written explanations for suggested exercises, which were performed 3 times daily and were graduated from week to week. Exercises were demonstrated to patients prior to hospital discharge. Patients in the unsupervised home exercise group were evaluated 2 weeks postoperatively, and those who were deemed to be behind in their recovery or who wished to attend formal outpatient therapy were allowed to do so. All participants from both groups were provided a diary to keep a record of their daily therapy regimen and to promote compliance. Additionally, patients were instructed to use a walker for 1 to 2 weeks and a cane for 2 additional weeks regardless of their allocation in the study.
The primary outcome was change in the Harris hip score (HHS) from baseline assessed longitudinally at both 1 month and 6 to 12 months postoperatively. The HHS is a validated, objective clinical evaluation tool based on pain, function, and hip range of motion. This score ranges from 0 to 100 points, with a higher score indicating a better outcome. Secondary outcomes included the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and Short Form-36 Health Survey (SF-36). The WOMAC is a validated patient-reported outcome tool that records pain, stiffness, and function. The inverse score is obtained by deducting subscale scores from a maximum of 100 points, with a higher score indicating a better outcome. Finally, the SF-36 is a measure of general health reported independently as mental and physical components. Both component scales are normalized to the U.S. population along a 0-to-100 range (mean of 50 points), with a higher score indicating less disability. Functional outcome instruments were administered in the office preoperatively, at the 1-month postoperative visit, and at a second routine visit that occurred at 6 to 12 months postoperatively.
This study was powered as a noninferiority trial contingent on a 2-sided p < 0.05 for significance and a power level of 0.80. Using the commonly reported standard deviation of 13 points and a noninferiority margin based on the minimum clinically important difference of 7 points for the HHS, which provided an effect size of 0.54, the necessary sample size was determined to be 55 patients per group20. Allowing for 10% attrition, 120 patients were recruited.
Baseline characteristics were compared between the groups; continuous variables were analyzed using the Mann-Whitney U test and categorical variables, using the Fisher exact test. The primary analysis of outcomes for this trial was conducted on an intention-to-treat basis, in that patients were analyzed based on their group allocation and adherence was ignored. A linear mixed-effect model taking into account longitudinal repeated measures and adjusting for potential confounders (age, sex, BMI, comorbidity, surgical approach) was used to assess differences between the groups. Significance was set at p < 0.05 for 2-sided tests, and all analyses were performed using R Statistical Computing Environment (R Foundation).
We identified 640 potentially eligible patients; 520 patients were not randomized, as they did not meet the inclusion criteria, could not be contacted, or declined to participate (Fig. 1). In the study, 120 patients consented and were enrolled. Twelve of these patients were excluded, as they opted out of the study, failed to follow up, or chose not to undergo the surgical procedure, leaving 108 patients (54 in the formal outpatient physical therapy group and 54 in the unsupervised home exercise group) in the final cohort. A total of 30 patients (28%) crossed over between groups: 20 (37%) from the formal outpatient physical therapy group and 10 (19%) from the unsupervised home exercise group.
Patient Demographic Characteristics
For patients included in the analysis, the mean age was 61.7 years (range, 38.8 to 79.6 years), the mean BMI was 29.3 kg/m2 (range, 19 to 42.9 kg/m2), the mean length of stay was 1.14 days (range, 0 to 3 days), the mean unadjusted Charlson Comorbidity Index was 0.269 (range, 0 to 4), the mean age-adjusted Charlson Comorbidity Index was 2.89 (range, 0 to 7), and the mean American Society of Anesthesiologists (ASA) score was 2.40 (range, 1 to 3). The cohort consisted of 61 male patients (56%). Eighty-six patients (80%) underwent total hip arthroplasty using the direct lateral approach. There was no significant difference (p > 0.05) in age, sex, length of hospital stay, comorbidity, surgical approach, or baseline function between the 2 groups. BMI was slightly greater in the formal outpatient therapy group (p = 0.04) (Table I). Patients lost or excluded from the study (n = 12) were considerably younger (52.4 years) than their counterparts (61.7 years) (p = 0.012).
Primary and Secondary Outcomes
Patients in both intention-to-treat groups had a significant improvement in function as measured by all administered instruments (p < 0.0001 for all outcomes), although the mental health component of the SF-12 did not change appreciably (p = 0.70) (Figs. 2 and 3). Improvement in the primary outcome, the HHS, from preoperative baseline to the first postoperative visit at 1 month was 21.5 points (95% confidence interval [CI], 16.2 to 26.9 points) for the formal outpatient therapy cohort and 23.3 points (95% CI, 18.3 to 28.4 points) for the unsupervised home exercise cohort. At the 6 to 12-month follow-up, patients in the formal outpatient therapy cohort had improved by 36.0 points (95% CI, 30.9 to 41.2 points) and those in the unsupervised home exercise cohort had improved by 35.6 points (95% CI, 30.9 to 40.4 points). The difference between the groups was minimal at both 1 month and 6 to 12 months postoperatively, and there was no significant difference in the HHS when controlling for confounders (p = 0.82) (Figs. 3 and 4).
Similar to the primary outcome, WOMAC scores had improved by 36.9 points (95% CI, 32.2 to 41.8 points) for the formal outpatient therapy cohort and 36.4 points (95% CI, 31.8 to 41.1 points) for the unsupervised home exercise cohort from baseline to the first postoperative month, and by 48.3 points (95% CI, 43.6 to 53.0 points) for the formal outpatient therapy cohort and 47.6 points (95% CI, 43.2 to 52.0 points) for the unsupervised home exercise cohort from baseline to 6 to 12 months. The difference in improvement between the groups at 1 month and 6 to 12 months was not found to be different (adjusted p = 0.80). Following the same general trend, SF-36 physical health scores had improved by 10.1 points (95% CI, 6.7 to 13.7 points) for the formal outpatient therapy cohort and 10.7 points (95% CI, 7.4 to 14.2 points) for the unsupervised home exercise cohort from baseline to the first postoperative month, and by 20.4 points (95% CI, 17.2 to 23.7 points) for the formal outpatient therapy cohort and 19.9 points (95% CI, 16.8 to 23.0 points) for the unsupervised home exercise cohort from baseline to 6 to 12 months. Again, there was no significant difference between groups (adjusted p = 0.90). Finally, SF-36 mental health scores did not improve significantly (p = 0.70) after total hip arthroplasty; the mental health component improved by 1.6 points (95% CI, −1.7 to 4.8 points) for the formal outpatient therapy group and 0.43 point (95% CI, −2.7 to 3.59 points) for the unsupervised home exercise group. Likewise, there was no difference between the 2 groups at 1 month or 6 to 12 months postoperatively (adjusted p = 0.34).
In this study, 28% of patients crossed over between the groups. Ten patients who chose to switch into formal outpatient therapy were older (66.2 years in the crossover group compared with 61.4 years in the per-protocol group; p = 0.08) (Table II). The small number of patients (10) who crossed over precluded formal analysis of outcomes measures in this cohort. In addition, 20 patients chose to switch out of formal outpatient therapy in favor of unsupervised home exercise. There was no difference in demographic characteristics or baseline function for this group (Table II).
It is well known that a large majority of patients undergoing total hip arthroplasty regain function and experience relief of pain. Numerous improvements in both surgical and anesthesia techniques, including minimized soft-tissue dissection, modern implants requiring modest intramedullary instrumentation, blood conservation, multimodal analgesia, and hypotensive regional anesthesia, allow patients undergoing total hip arthroplasty to walk early and with minimal pain. With these strides, informal home exercise may be sufficient for most patients at a substantial cost savings. Post-discharge costs account for as much as 40% of a total hip arthroplasty episode, in large part because of rehabilitation services14,21,22. As such, the value of formal physical therapy in the modern era has been raised as a practical clinical question14,15,23. In this randomized clinical trial, we aimed to determine whether unsupervised home exercise after total hip arthroplasty is similarly effective compared with formal outpatient therapy, which would justify a shift in routine care.
The results of this study strongly reinforced the growing perception that informal home exercise can provide the same improvement in function and quality of life as that of formal outpatient therapy. Substantial postoperative improvement in all recorded outcomes was demonstrated for both rehabilitation modalities. We found no difference between the groups to be significant or, more importantly, clinically important. Intangible differences at various time points were well below the minimum that would be perceived clinically as beneficial by patients24.
This major finding has led all joint surgeons at our institution to discontinue the routine use of formal outpatient therapy for most patients after total hip arthroplasty. However, some patients may still benefit from formal outpatient therapy. Although the current study did not aim to identify a subset of patients who may still require supervised rehabilitation, 10 patients partaking in self-directed home exercise did not meet progress goals after 2 weeks and were transitioned into formal outpatient therapy. These patients tended to be older and had worse preoperative function. It is well established that poor preoperative functional status is associated with worse postoperative functional outcomes, and such patients may be more likely to benefit from formal outpatient therapy25-28. Additional indications for referral to formal outpatient therapy may include severe preoperative gait imbalance, postoperative neurological complications, or potentially a patient expectation for a quicker return to high-level physical activity23.
We also examined a subset of 20 patients who chose to forgo formal outpatient therapy to which they were assigned in favor of unsupervised home exercise. Most patients cited out-of-pocket expenses and logistical constraints as substantial burdens. In our cohort, non-Medicare patients were responsible for copayments ranging from $10 to $60 per session, which can amount to as much as $1,440 for 8 weeks of physical therapy. Patients also found it difficult to take time off from work to attend therapy sessions or to find transportation to physical therapy facilities, particularly during the first few weeks when they could not drive. Instead, these patients found prescribed home exercise to be much more convenient and practical.
One must also consider the cost to the health-care system when deciding on a clinical intervention. At a mean cost of $2,500 per episode, a 10-week formal physical therapy regimen for 54 patients in the formal outpatient physical therapy group had an estimated cost of $135,00014. This can be compared with a cost of $20,000 for 10 patients in the unsupervised home exercise group who required outpatient therapy services after 2 weeks, which represents an estimated cost savings of $115,000 ($2,130 per patient). If the cost savings from an unsupervised home exercise program were applied to all 326,100 patients who underwent total hip arthroplasty in the United States in 2010, the potential cost savings would have been nearly $695 million29.
This study had several notable limitations, which may have restricted our ability to detect an advantage for formal outpatient therapy. As with many randomized clinical trials, a substantial number of patients were noncompliant with the allocated intervention. However, the intention-to-treat strategy is a widely accepted approach for overcoming this common flaw. An intention-to-treat analysis maintains the prognostic balance of the original randomization, as it reflects the real-life clinical scenario and avoids overly optimistic estimates of treatment effect. This strategy also avoids systematic exclusion of the noncompliant patient cohort; noncompliance may, in fact, reflect response to treatment. Still, critics suggest that the estimate of treatment effect is too conservative with this strategy, which could increase the risk for committing a type-II statistical error for studies of a noninferiority nature30. Additionally, total hip arthroplasty was performed using 2 different surgical approaches, which could be deemed a limitation. However, perioperative and postoperative protocols were not influenced by surgical approach, minimizing the potential for bias. No cases were performed using a posterolateral approach, and thus the validity of our conclusions for surgeons utilizing this surgical approach remains unclear.
In addition, selection bias may have influenced our cohort and even may have impacted the course of treatment for some patients. Patients of a higher socioeconomic and education level may have had a greater appreciation for the value of the study and thus more commonly may have agreed to participate. These same patients also may be more motivated to follow through, performing assigned exercises with greater frequency and accuracy. Moreover, patients perceiving themselves to be at a heightened risk for a poor outcome may have elected not to participate.
Lastly, the HHS, WOMAC, and SF-36 all have been reported to have floor and ceiling effects, in that a high proportion of respondents may grade themselves as having a maximal or minimal score for that instrument24,31. Such phenomena limit the aptitude of outcome measures in perceiving more subtle differences between patients, especially for interventions that result in generally good or poor outcomes.
Despite the aforementioned limitations, this study provides evidence to support the equivalence of unsupervised home exercise as an effective rehabilitation strategy for most patients, justifying its use as a standard of routine care after total hip arthroplasty. However, as mentioned above, we must emphasize that certain patient populations may require formal physical therapy to optimize their recovery after total hip arthroplasty.
NOTE: The authors are grateful to Mitchell Maltenfort for his essential contribution of all statistical analyses in this manuscript.
Investigation performed at The Rothman Institute, Department of Orthopaedic Surgery, Thomas Jefferson University, Philadelphia, Pennsylvania
Disclosure: This project did not receive any financial funding from external sources. 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 relevant financial relationship in the biomedical arena outside the submitted work (http://links.lww.com/JBJS/C257).
1. Ethgen O, Bruyère O, Richy F, Dardennes C, Reginster JY. Health-related quality of life in total hip and total knee arthroplasty. A qualitative and systematic review of the literature. J Bone Joint Surg Am. 2004 ;86(5):963–74.
2. Ng CY, Ballantyne JA, Brenkel IJ. Quality of life and functional outcome after primary total hip replacement. A five-year follow-up. J Bone Joint Surg Br. 2007 ;89(7):868–73.
3. Nilsdotter AK, Petersson IF, Roos EM, Lohmander LS. Predictors of patient relevant outcome after total hip replacement for osteoarthritis: a prospective study. Ann Rheum Dis. 2003 ;62(10):923–30.
4. Beswick AD, Wylde V, Gooberman-Hill R, Blom A, Dieppe P. What proportion of patients report long-term pain after total hip or knee replacement for osteoarthritis? A systematic review of prospective studies in unselected patients. BMJ Open. 2012 ;2(1):e000435.
5. Galea MP, Levinger P, Lythgo N, Cimoli C, Weller R, Tully E, McMeeken J, Westh R. A targeted home- and center-based exercise program for people after total hip replacement: a randomized clinical trial. Arch Phys Med Rehabil. 2008 ;89(8):1442–7. Epub 2008 Jun 30.
6. Trudelle-Jackson E, Smith SS. Effects of a late-phase exercise program after total hip arthroplasty: a randomized controlled trial. Arch Phys Med Rehabil. 2004 ;85(7):1056–62.
7. Di Monaco M, Castiglioni C. Which type of exercise therapy is effective after hip arthroplasty? A systematic review of randomized controlled trials. Eur J Phys Rehabil Med. 2013 ;49(6):893–907; quiz 921-3. Epub 2013 Oct 30.
8. Coulter CL, Scarvell JM, Neeman TM, Smith PN. Physiotherapist-directed rehabilitation exercises in the outpatient or home setting improve strength, gait speed and cadence after elective total hip replacement: a systematic review. J Physiother. 2013 ;59(4):219–26.
9. Umpierres CS, Ribeiro TA, Marchisio ÂE, Galvão L, Borges ÍN, Macedo CA, Galia CR. Rehabilitation following total hip arthroplasty evaluation over short follow-up time: randomized clinical trial. J Rehabil Res Dev. 2014;51(10):1567–78.
10. Husby VS, Helgerud J, Bjørgen S, Husby OS, Benum P, Hoff J. Early postoperative maximal strength training improves work efficiency 6-12 months after osteoarthritis-induced total hip arthroplasty in patients younger than 60 years. Am J Phys Med Rehabil. 2010 ;89(4):304–14.
11. Suetta C, Andersen JL, Dalgas U, Berget J, Koskinen S, Aagaard P, Magnusson SP, Kjaer M. Resistance training induces qualitative changes in muscle morphology, muscle architecture, and muscle function in elderly postoperative patients. J Appl Physiol (1985). 2008 ;105(1):180–6. Epub 2008 Apr 17.
12. Lowe CJ, Davies L, Sackley CM, Barker KL. Effectiveness of land-based physiotherapy exercise following hospital discharge following hip arthroplasty for osteoarthritis: an updated systematic review. Physiotherapy. 2015 ;101(3):252–65. Epub 2015 Jan 16.
13. Mikkelsen LR, Mechlenburg I, Søballe K, Jørgensen LB, Mikkelsen S, Bandholm T, Petersen AK. Effect of early supervised progressive resistance training compared to unsupervised home-based exercise after fast-track total hip replacement applied to patients with preoperative functional limitations. A single-blinded randomised controlled trial. Osteoarthritis Cartilage. 2014 ;22(12):2051–8. Epub 2014 Oct 22.
14. Ong KL, Lotke PA, Lau E, Manley MT, Kurtz SM. Prevalence and costs of rehabilitation and physical therapy after primary TJA. J Arthroplasty. 2015 ;30(7):1121–6. Epub 2015 Feb 28.
15. Minns Lowe CJ, Barker KL, Dewey ME, Sackley CM. Effectiveness of physiotherapy exercise following hip arthroplasty for osteoarthritis: a systematic review of clinical trials. BMC Musculoskelet Disord. 2009 ;10:98.
16. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol. 1994 ;47(11):1245–51.
17. Restrepo C, Parvizi J, Pour AE, Hozack WJ. Prospective randomized study of two surgical approaches for total hip arthroplasty. J Arthroplasty. 2010 ;25(5):671–9.e1. Epub 2010 Apr 8.
18. Austin MS, Rothman RH. Acetabular orientation: anterolateral approach in the supine position. Clin Orthop Relat Res. 2009 ;467(1):112–8. Epub 2008 Oct 1.
19. Peak EL, Parvizi J, Ciminiello M, Purtill JJ, Sharkey PF, Hozack WJ, Rothman RH. The role of patient restrictions in reducing the prevalence of early dislocation following total hip arthroplasty. A randomized, prospective study. J Bone Joint Surg Am. 2005 ;87(2):247–53.
20. Kalairajah Y, Azurza K, Hulme C, Molloy S, Drabu KJ. Health outcome measures in the evaluation of total hip arthroplasties—a comparison between the Harris hip score and the Oxford hip score. J Arthroplasty. 2005 ;20(8):1037–41.
21. Lavernia CJ, Hernandez VH, Rossi MD. Payment analysis of total hip replacement. Curr Opin Orthop. 2007;18:23–7.
22. Bozic KJ, Ward L, Vail TP, Maze M. Bundled payments in total joint arthroplasty: targeting opportunities for quality improvement and cost reduction. Clin Orthop Relat Res. 2014 ;472(1):188–93. Epub 2013 May 7.
23. Artz N, Dixon S, Wylde V, Beswick A, Blom A, Gooberman-Hill R. Physiotherapy provision following discharge after total hip and total knee replacement: a survey of current practice at high-volume NHS hospitals in England and Wales. Musculoskeletal Care. 2013 ;11(1):31–8. Epub 2012 Jul 9.
24. Quintana JM, Escobar A, Bilbao A, Arostequi I, Lafuente I, Vidaurreta I. Responsiveness and clinically important differences for the WOMAC and SF-36 after hip joint replacement. Osteoarthritis Cartilage. 2005 ;13(12):1076–83. Epub 2005 Sep 9.
25. Kennedy DM, Hanna SE, Stratford PW, Wessel J, Gollish JD. Preoperative function and gender predict pattern of functional recovery after hip and knee arthroplasty. J Arthroplasty. 2006 ;21(4):559–66.
26. Holtzman J, Saleh K, Kane R. Effect of baseline functional status and pain on outcomes of total hip arthroplasty. J Bone Joint Surg Am. 2002 ;84(11):1942–8.
27. Jensen C, Aagaard P, Overgaard S. Recovery in mechanical muscle strength following resurfacing vs standard total hip arthroplasty – a randomised clinical trial. Osteoarthritis Cartilage. 2011 ;19(9):1108–16. Epub 2011 Jun 28.
28. Fortin PR, Clarke AE, Joseph L, Liang MH, Tanzer M, Ferland D, Phillips C, Partridge AJ, Bélisle P, Fossell AH, Mahomed N, Sledge CB, Katz JN. Outcomes of total hip and knee replacement: preoperative functional status predicts outcomes at six months after surgery. Arthritis Rheum. 1999 ;42(8):1722–8.
29. Wolford ML, Palso K, Bercovitz A. Hospitalization for total hip replacement among inpatients aged 45 and over: United States, 2000-2010. NCHS Data Brief. 2015 ;(186):1-8.
30. Gupta SK. Intention-to-treat concept: a review. Perspect Clin Res. 2011 ;2(3):109–12.
31. Wamper KE, Sierevelt IN, Poolman RW, Bhandari M, Haverkamp D. The Harris hip score: do ceiling effects limit its usefulness in orthopedics? Acta Orthop. 2010 ;81(6):703–7.