|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 19 (July 1), 2005: pp. 4414-4423 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.01.040 Persistence of Medical Change at Implementation of Clinical Guidelines on Medical Practice: A Controlled Study in a Cancer NetworkFrom the Centre Léon Bérard; Groupe de Recherche en Economie de la Santé et Réseaux de Soins en Cancérologie, Lyon; Clinique La Digonnière, Saint-Etienne; Centre Hospitalier de Chambéry, Chambéry; Centre Hospitalier de Roanne, Roanne; Clinique de Rillieux, Rillieux; Centre Val d'Aurelle; Clinique Clementville, Montpellier; and Fédération Nationale des Centres de Lutte Contre le Cancer, Paris, France Address reprint requests to Isabelle Ray-Coquard, MD, Centre Léon Bérard, 28, rue Laënnec, 69008 Lyon, France; e-mail: ray{at}lyon.fnclcc.fr
PURPOSE: A cancer network of general or private hospitals of a French region was started in 1995 for improving quality of care and rationalizing medical prescriptions. The impact of implementing a clinical practice guidelines (CPG) project assessed conformity with guidelines in medical practice; significant changes were observed within the network, whereas no changes were observed in a control region without cancer network. In the present study, we evaluated the persistence of conformity to guidelines through a new medical audit. PATIENTS AND METHODS: In 1999, the hospitals of the previously compared experimental and control groups accepted to reassess the impact of CPG. A controlled transversal study was performed in the experimental group (cancer network) and in the control group (no regional cancer network). In 1996 (first audit) and in 1999 (present audit), all new patients with colon cancer (177 and 200 in experimental group and 118 and 100 in control group, respectively) and early breast cancer (444 and 381 in experimental group and 172 and 204 in control group, respectively) were selected. RESULTS: In the experimental group, the compliance of medical decisions with CPG was significantly higher in 1999 than in 1996 for colon cancer (73%; 95% CI, 67% to 79% v 56%; 95% CI, 49% to 63%, respectively; P = .003) and similar for the two periods for breast cancer (36%; 95% CI, 31% to 41% v 40%; 95% CI, 35% to 44%, respectively; P = .24). In the control group, compliance was significantly higher in 1999 than in 1996 for colon cancer (67%; 95% CI, 58% to 76% v 38%; 95% CI, 29% to 47%, respectively; P < .001) and identical for the two periods for breast cancer (4%; 95% CI, 1% to 7% v 7%; 95% CI, 3% to 11%, respectively; P = .19). CONCLUSION: The CPG program for cancer management produced persistent changes in medical practice in our cancer network in terms of conformity with CPG.
There is increasing pressure on the medical profession to practice evidence-based medicine and to introduce measures of quality assurance applicable to the care of patients. Initiatives to improve the consistency between scientific evidence and clinical practice are increasingly focusing on managed care plans and integrated delivery systems. Studies documenting inappropriate use and great variations in practice have drawn attention to the possibility of improving health care quality and have raised hopes that cost containment and quality improvement can go hand in hand.1,2 Over the past decade, these factors have driven researchers' efforts to synthesize evidence about managing medical conditions.3 Implementing the resulting guidelines has been more difficult because it requires widespread behavior change.4,5 It has been shown that the simple distribution of guidelines and published reports from consensus conferences fail to modify physicians' practices.6 The successful introduction of clinical guidelines leading to significant improvements in clinical care depends on many factors including the clinical setting, the methods of development and dissemination, and, more importantly, the methods of implementation of the guidelines.7 We previously reported the successful development and implementation of clinical practice guidelines (CPG) for medical practice in one comprehensive cancer center in Lyon, France.8 Concomitantly, the cancer center provided leadership for cancer treatment in the Rhône-Alpes region and developed a managed care network model similar to those previously described.9 This approach was initiated in 1995, when hospitals located in the Rhône-Alpes region of France were invited to participate in the elaboration of a regional cancer network, Oncology Rhône-Alpes (ONCORA). The cancer network development plan involved the implementation of the CPG elaborated in the cancer center. In 2001, we conducted a controlled transversal study to evaluate the impact of the CPG and the implementation strategy applied in the cancer network (experimental group). In that study, an increased number of medical decisions compliant with the CPG was reported for the experimental group, whereas no change was noted in another region without cancer network organization (control group).10 However, it was well demonstrated that the stability of this level of compliance could not be maintained over time, especially after the intensive implementation strategy has stopped.11-14 In the present study, we planned to examine the rate of compliance with CPG over time. We will also assess whether the diffusion of presented results (feedback) can modify physicians' medical practice because this has not always been reported to be the case.15 An external medical record reaudit process was set up in the two previously studied regions to reassess the impact of CPG on diagnosis and treatment practice for breast and colon cancer patients; Rhône-Alpes, with its regional cancer network organization, was the experimental group. The previous anonymous matched control region, where neither regional cancer network nor process of guidelines implementation existed at the time of data collection, was the control group.
Objectives The question in this article was to investigate whether the network organization was capable of maintaining, after the implementation phase, good conformity and homogeneity of medical practice with guidelines. We reassessed the impact of CPG on the management of breast and colon cancers in the regional cancer network by comparing the compliance of medical practice with the CPG between 1996 and 1999 in the two matched groups of hospitals. As for the previous study, compliance in the two groups was compared between years of evaluation, not between groups of institutions, in so far as any CPG or cancer network organization was delineated in the control group. No statistical comparisons were made between groups because this was not the purpose of the study. When medical decisions were judged to be noncompliant, we verified whether they were based on scientific evidence from published literature.
Setting Control group. The control region (27,000 km2 and 2.3 million inhabitants) has approximately 80 private and 40 public hospitals (including three teaching hospitals), providing a total of about 11,000 beds. Approximately 9,000 new malignant tumors are treated annually in the region.
Intervention: Cancer Network Development and Implementation Strategy of the CPG
Study Design
Criteria for Hospital, Region, and Tumor Selection Regions. The experimental group was the Rhône-Alpes region, where the CPG program was implemented; it belongs to the French regional cancer network ONCORA. The control group was chosen among regions that did not participate in a CPG program, in a cancer network, or in the ONCORA regional cancer network organization. The control region was identical to that of the previous study; it was chosen for reasons of similarity with the control group with respect to the existence of medical school training and teaching hospitals (although teaching hospitals were excluded from the study), a mix of private and public institutions, and over 5,000 newly diagnosed cancers reported annually.10 Institutions. Each group comprised the same institutions as previously evaluated, from both private and public sectors.10 To qualify for the experimental group, hospitals of the Rhône-Alpes region had to be participating in regional cancer network activities since 1995. Tumors. We chose to study localized breast cancer and colon cancers (any stage) because both tumors are representative of medical practice in oncology and in the network (frequency, numerous sources of evidence for breast cancer and few for colon cancer).
Selection of Patients' Records
Data Sources
Measurements
Regarding breast cancer, only the initial overall treatment sequence, including surgery, radiotherapy, chemotherapy, and hormone therapy with initial examination and follow-up (when these were performed in the participating hospital), was taken into account. Therefore, only one overall treatment sequence could be assessed for each patient. Regarding colon cancer, data of both initial treatment and treatment for relapses were considered because CPG were available for all phases of disease management; hence, several overall treatment sequences could be assessed for colon cancer patients. The second level of conformity was the number of medical decisions judged to be based on either the CPG or published randomized trials (ie, evidence based). This was assessed by a systematic literature search using a documented strategy as developed by Guyatt et al.18 All types of procedures were individually classified as being either evidence based if they complied with CPG recommendations or based on scientific evidence (scientific-level graded from I to II18) if established in one or more randomized controlled trial or overviews of randomized controlled trials. When no published randomized trials, uncontrolled studies, or national guidelines recommendations19 could be identified for at least one individual medical decision, the overall treatment sequence was classified as displaying no convincing supporting scientific evidence.
Statistical Analysis
Breast Cancer As reported earlier, only one overall treatment sequence could be assessed for each patient. Patient characteristics. In both groups of patients, some records were not assessable for certain types of procedures individually (Table 1). No statistical difference was noted between patient characteristics in 1996 and 1999 when comparing patients of both the experimental and control groups (comparison not shown). For both groups of hospitals, the characteristics of the patients were similar in 1996 and 1999 regarding age, tumor size, and number of involved and examined nodes; however, hormone receptor status and Scarf Bloom Richardson score differed significantly between the two periods in the two groups. Less than 5% of data were missing for the two groups of patients in 1999 (Table 2).
Compliance with CPG. In the experimental group, the observed rates of compliance with CPG of the 825 assessable overall treatment sequences were similar for both periods, with a rate of 36% in 1999 (137 of 381 treatment sequences) and 40% in 1996 (178 of 444 treatment sequences; P = .25). In addition, the number of individual medical decisions complying with the CPG was also significantly higher in 1999 than in 1996 for hormone therapy, whereas surgery procedures were statistically less compliant in 1999 than in 1996 (Table 3).
In the control group, we detected no difference in the observed rate of compliance with CPG of the 376 assessable overall treatment sequences in 1996 (7%; 12 of 172 treatment sequences) and in 1999 (4%; eight of 204 treatment sequences; P = .99). Of note, compliance of hormone therapy and radiation therapy procedures was significantly higher in 1999 than in 1996. Also, in the control group, surgical procedures were statistically less compliant in 1999 compared with 1996 (Table 3). Stratified analysis by center in both groups showed that the numbers of overall treatment sequences complying with the CPG in each hospital of the experimental and control groups were similar in 1999 and 1996 (data not shown). Of note, only the cancer centers maintained their initial compliance rate for surgical procedure (> 85% and 75% in the experimental and control groups, respectively), whereas compliance rate decreased to less than 70% in all other institutions. Evidence-based medical decisions. When medical decisions were judged to be noncompliant according to CPG, we verified whether they were based on scientific evidence from published literature. For the experimental group, in 1996 and 1999, 50% (95% CI, 45% to 55%; 223 of 444 treatments) and 44% (95% CI, 39% to 49%; 166 of 381 treatments; P = .01) of initial treatments were consistent with the CPG or were judged to be based on scientific evidence, respectively. In the control group, these results were 14 of 172 treatments in 1996 (8%; 95% CI, 4% to 12%) and 20 of 200 treatments in 1999 (10%; 95% CI, 6% to 14%; P = .58). Decisions with no convincing scientific evidence concerned essentially, with decreasing frequencies, the following: (1) patient surveillance after treatment and too intensive follow-up compared with the simple locoregional surveillance (clinical evaluation every 4 months and annual mammography examination) supported by the literature as recommended in the CPG (76 and 308 patients in the experimental and control groups, respectively); (2) less than 10 nodes in the axillary dissection for surgery procedure (81 and 54 patients in the experimental and control groups, respectively); (3) the use of nonvalidated chemotherapy regimens in the adjuvant phase or absence of chemotherapy for patients with positive nodes (18 and 37 patients in the experimental and control groups, respectively); and (4) the lack of adjuvant treatment for patients with positive hormone receptors (three and 10 patients in the experimental and control groups, respectively).
Colon Cancer No guidelines concerning initial examination and follow-up for metastatic disease have been set up in the CPG; thus, only a few medical decisions were assessable for these aspects of practice. For the experimental and control groups, only 106 and 94 records, respectively, were assessable for follow-up because follow-up was performed elsewhere (n = 53), because the patient had died (n = 23), or because of early relapse at the time of analysis (n = 29). Patient characteristics. No statistical difference was observed between patient characteristics in 1996 and 1999 when comparing patients in the experimental group and patients in the control group (data not shown). Also, for both groups of hospitals, the characteristics of the patients were similar in 1996 and 1999 (Table 4).23
Compliance with CPG. In the experimental group, the observed rate of compliance with CPG of assessable overall treatment sequences was significantly higher in 1999 (73%; 161 of 221 treatment sequences) than in 1996 (56%; 118 of 211 treatment sequences; P = .03; Table 3). In addition, the number of individual medical decisions complying with the CPG was also significantly higher in 1999 than in 1996, except for surgical procedures and follow-up. In the control group, the observed rate of compliance with CPG of assessable overall treatment sequences was significantly higher in 1999 (67%; 67 of 100 treatment sequences) than in 1996 (38%; 48 of 125 treatment sequences; P < .001). In addition, the number of individual medical decisions complying with the CPG was also significantly higher in 1999 than in 1996, except for initial examination and follow-up. Stratified analysis in both groups showed that two hospitals of the experimental group had significantly higher numbers of overall treatment sequences complying with the CPG in 1999 than in 1996, whereas two hospitals remained stable (data not shown). No change was noted for one hospital of the control group, whereas two hospitals significantly increased their rate of compliance with CPG in 1999 compared with 1996 (data not shown).
Because both local and metastatic diseases were evaluated in colon cancer, a Mantel-Haenszel Evidence-based medical decisions. For the experimental group, in 1999 and 1996, 75% (95% CI, 69% to 81%; 150 of 200 treatments) and 83% (95% CI, 76% to 89%; 176 of 211 treatments) of initial treatments, respectively, conformed with the CPG or were judged to be based on scientific evidence (P = .49). In the control group, these results were 74 of 125 treatments in 1996 (59%; 95% CI, 50% to 67%) and 68 of 104 treatments in 1999 (68%; 95% CI, 59% to 77%; P = .01). Decisions with no convincing scientific evidence concerned essentially, with decreasing frequencies, the following: (1) nonscientifically based chemotherapy procedures, including the lack of adjuvant or palliative chemotherapy as well as the prescription of nonvalidated chemotherapy regimens (63, all in 1996, and 48 patients in the experimental and control groups, respectively); (2) the lack of colonoscopy control after colon surgery for patient follow-up (eight and seven patients in the experimental and control groups, respectively); and (3) the administration of adjuvant radiation therapy (two and five patients in the experimental and control groups, respectively).
The present study indicates that, in this network, CPG were able to produce sustained improvements in adherence to medical practice over time compared with a control region. It has been reported that women treated according to consensus recommendations for systemic therapy experience a significant improvement in survival.24 This prompted us to develop methods capable of enhancing the level of conformity to CPG. In reauditing medical decisions of the cancer network some years after the original audit, our aim was to measure the level of conformity to guidelines and to detect a possible decrease of guidelines use by clinicians.10 No decrease was observed after a 3-year period. Conversely, an increase or stabilization was found in the experimental group for both breast and colon treatments. These results suggest that the implementation of the CPG in the network has sustained efficacy. To our knowledge, this is the first report showing that the development and implementation strategy of the CPG program for cancer management may produce persistent changes in medical practice in a cancer network in term of conformity with CPG. Actually, recently published studies reported the absence of significant improvement in mean waiting time compared with previous audits.25-27 Results observed in hospitals of this experimental group could illustrate the fact that implementing new organizational strategies, such as cancer networks, is an option for durably changing clinical practice; however, improvements were not obtained for all medical procedures.28 Most studies have noted the importance of involving physicians in the process of quality improvement. Interventions based on group dynamics and sensitive to local practice have been shown to be useful in facilitating the adoption of guidelines by physicians. Well- designated studies, especially those regarding medications, have obtained significant improvements from relying on opinion leaders or face-to-face persuasion by detailers.29,30 As was the case in our experimental cancer network, most organizations (health maintenance organizations, national medical societies, and so on) have attained greater success associated with elements of feedback, including participation of clinical leaders, targeting specific problems with explicit standards, information that physicians can act on, discussion in small groups, and repeated feedback.9 However, results reported in this study and in all previous studies can still be significantly improved. Although compliance rates in the experimental group were better than in the control group regarding CPG and evidence-based medicine, percentages were generally less than 70% for the overall treatment sequence and 80% for each procedure. Indeed, the opinion leader approach implies that the appropriate clinician will vary depending on the clinical topic (eg, the choice in chemotherapy regimen for breast cancer v colon cancer). From that perspective, establishing a routine, ongoing process under management auspices that deal with many clinical issues may prove more effective than relying on opinion leaders or detailing alternatives.31,32 A possible bias of repeating measures is the evolution of the CPG from one period to the other. The use and spreading of CPG make them more and more detailed and complex over time to take into account more different medical situations. Consequently, the risk of inappropriate decisions increases with time. The problem of compliance with the surgical procedure for node dissection in both diseases illustrates well the difficulties in reauditing medical procedures when the CPG has changed. Indeed, the CPG in the cancer network became more and more detailed over time. The first reason for noncompliance was the low number (< 10 nodes) of examined nodes in the axillary dissection compared with the minimum of 10 nodes per dissection recommended by the CPG. This recommendation was made only after the previous version of the CPG16 had been used for the 2001 evaluation.10 Similarly, as in breast cancer, surgery for colon cancer was found to be significantly less compliant with the CPG in the control group because of the noncompliance with the requirement of more than eight nodes retrieved on resection after lymphadenectomy in colon cancer surgery.33-35 Another example is that, in 1995, the CPG simply recommended radiation therapy after tumorectomy for localized breast cancer, whereas, in 1996, radiation therapy was specified in term of fields and, in 1999, in term of dose. These details show that medical procedures used in 1999 were judged against recommendations that were more complex and refined over time. This increase in precision and complexity is indeed partially responsible for modifications to the compliance rate. The same comments apply to medical decisions based on scientific evidence. In this new study, medical practice is frequently described as having no convincing supporting scientific evidence, essentially because guidelines, consensus, and scientific publications of 2001 more frequently described inadequate management, notably for localized breast cancer.19,36,37 These observations reinforced the necessity of a matched control group so that results are not misinterpreted either positively or negatively. Limitations also concerned the inability to measure other external or internal influences possibly accounting for some of the changes between 1996 and 1999. For instance, we could not evaluate the potential impact of each institution's policy on both control and experimental intervention groups. It would be interesting to test the possible contribution of specific institutions' policies in term of exposure to guidelines or other educational systems and to measure the effect of this exposure. Of equal importance is the need to understand large disparities in medical practice within geographic regions. The other important observation that emerged from this study is the considerable financial and human cost of this type of evaluation. For the final step developed over the last year, one technician and two medical doctors were requested for elaborating, developing, and achieving result evaluation and analysis. We cannot presently recommend this type of evaluation as being an easy procedure to evaluate medical practice daily. It would be unrealistic to suggest that hospitals use this type of medical audit, involving a control group and controls in the scientific literature of medical decisions not covered by the CPG, for their annual audit of good practice. Results of this and other studies8,10 prompt us to delineate factors of substitution valid for more easy and realistic hospital evaluation after a possible first complete analysis. Possible surrogate markers, involving parameters related to structures, practitioners, and patient characteristics, will be explored using the different data of the present and two previous studies. We plan to delineate, assess, and possibly validate these factors in view of facilitating future attempts by a given hospital to evaluate compliance to strategic factors. As the pressure mounts to practice evidence-based medicine, it will become more important for physicians and institutions to understand how to select appropriate guidelines, how to integrate guidelines into routine practice, and how to measure compliance and overall benefit to the patient when compliance is achieved. Results presented in this article show that adapting the external dissemination process makes it possible to extend the consensus for the CPG in a target audience and, thus, produce a behavior change.38 When applied to physicians in community hospitals, our educational strategy of opinion leaders performing detailing based on the CPG produced a significant behavior change. Results described for colon cancer management in this last step suggest that the behavior change was more rapidly obtained in the experimental group than in a region with no organizational network, suggesting that valid information could reach the target more rapidly.
The authors indicated no potential conflicts of interest.
We thank Marie Dominique Reynaud for editing assistance, Josiane Popescu for Medline search, and Florence Germaine for assistance in collecting the data.
Supported by ANAES and CNAM grants from the Ministry of Health (France). Presented in part at the 39th Annual Meeting of American Society Clinical Oncology, Chicago, IL, May 31-June 3, 2003. Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Lazovich DA, White E, Thomas DB, et al: Underutilization of breast-conserving surgery and radiation therapy among women with stage I or II breast cancer. JAMA 266:3433-3438, 1991 2. Farrow DC, Hunt WC, Samet JM: Geographic variation in the treatment of localized breast cancer. N Engl J Med 326:1097-1101, 1992[Abstract]
3. Guyatt GH, Sackett DL, Cook DJ: Users' guides to the medical literature: II. How to use an article about therapy or prevention—A. Are the results of the study valid? Evidence-Based Medicine Working Group. JAMA 270:2598-2601, 1993
4. Davis DA, Norman GR, Painvin A, et al: Attempting to ensure physician competence. JAMA 263:2041-2042, 1990 5. Ellis J, Mulligan I, Rowe J, et al: Inpatient general medicine is evidence based: A-Team, Nuffield Department of Clinical Medicine. Lancet 346:407-410, 1995[CrossRef][Medline] 6. Ferguson JH: NIH consensus conferences: Dissemination and impact. Ann N Y Acad Sci 703:180-198, 1993[CrossRef][Medline]
7. Eddy DM: Clinical decision making: From theory to practice. Practice policies—Guidelines for methods. JAMA 263:1839-1841, 1990
8. Ray-Coquard I, Philip T, Lehmann M, et al: Impact of a clinical guidelines program for breast and colon cancer in a French cancer center. JAMA 278:1591-1595, 1997 9. Sisk JE: How are health care organizations using clinical guidelines? Health Aff (Millwood) 17:91-109, 1998[Abstract] 10. Ray-Coquard I, Philip T, de Laroche G, et al: A controlled "before-after" study: Impact of a clinical guidelines programme and regional cancer network organization on medical practice. Br J Cancer 86:313-321, 2002[CrossRef][Medline] 11. Murray MD, Harris LE, Overhage JM, et al: Failure of computerized treatment suggestions to improve health outcomes of outpatients with uncomplicated hypertension: Results of a randomized controlled trial. Pharmacotherapy 24:324-337, 2004[CrossRef][Medline] 12. Tierney WM, Miller ME, McDonald CJ: The effect on test ordering of informing physicians of the charges for outpatient diagnostic tests. N Engl J Med 322:1499-1504, 1990[Abstract] 13. Tierney WM: Improving clinical decisions and outcomes with information: A review. Int J Med Inform 62:1-9, 2001[CrossRef][Medline] 14. Dexter PR, Wolinsky FD, Gramelspacher GP, et al: Opportunities for advance directives to influence acute medical care. J Clin Ethics 14:173-182, 2003[Medline] 15. Russell IT, Grimshaw JM: The effectiveness of referral guidelines: A review of the methods and findings of published evaluations, in Rolland M, Coulter A: Hospital Referral. Oxford, United Kingdom, Oxford University Press, 1992, pp 179-211 16. Centre Léon Bérard and Réseau ONCORA: Thesaurus ONCORA en Cancérologie. Paris, France, Arnette Blackwell, 1997 17. Donabedian A: Contributions of epidemiology to quality assessment and monitoring. Infect Control Hosp Epidemiol 11:117-121, 1990[Medline]
18. Guyatt GH, Sackett DL, Sinclair JC, et al: Users' guides to the medical literature: IX. A method for grading health care recommendations—Evidence-Based Medicine Working Group. JAMA 274:1800-1804, 1995 19. Fédération des Centres de Lutte Contre le Cancer: Cancers du Sein Infiltrants Non Métastatiques (vol 12, ed 2). Montrouge, France, Fédération des Centres de Lutte Contre le Cancer, 2001, pp 47-268 20. Mantel NHW: Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 22:719-748, 1959 21. Douillard JY, Cunningham D, Roth AD, et al: A randomized phase III trial comparing irinotecan (IRI) + 5FU/folinic acid (FA) to the same schedule of 5FU/FA in patients (pts) with metastatic colorectal cancer (MCRC) as front line chemotherapy (CT). Proc Am Soc Clin Oncol 18:233a, 1999 (abstr 899) 22. Douillard JY, Cunningham D, Roth AD, et al: Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer: A multicentre randomised trial. Lancet 355:1041-1047, 2000[CrossRef][Medline] 23. Astler VB, Coller FA: The prognostic significance of direct extension of cancer of the colon and rectum. Ann Surg 139:846-852, 1954[Medline]
24. Hebert-Croteau N, Brisson J, Latreille J, et al: Compliance with consensus recommendations for systemic therapy is associated with improved survival of women with node-negative breast cancer. J Clin Oncol 22:3685-3693, 2004 25. Chan SY, Berry MG, Engledow AH, et al: Audit of a one-stop breast clinic: Revisited. Breast Cancer 7:191-194, 2000[Medline] 26. Madhavan P, Blom A, Karagkevrakis B, et al: Deterioration of theatre discipline during total joint replacement: Have theatre protocols been abandoned? Ann R Coll Surg Engl 81:262-265, 1999
27. Duffett R, Lelliott P: Auditing electroconvulsive therapy: The third cycle. Br J Psychiatry 172:401-405, 1998
28. Wallace LM, Freeman T, Latham L, et al: Organisational strategies for changing clinical practice: How trusts are meeting the challenges of clinical governance. Qual Health Care 10:76-82, 2001
29. Davis DA, Thomson MA, Oxman AD, et al: Changing physician performance: A systematic review of the effect of continuing medical education strategies. JAMA 274:700-705, 1995
30. Lomas J, Enkin M, Anderson GM, et al: Opinion leaders vs audit and feedback to implement practice guidelines: Delivery after previous cesarean section. JAMA 265:2202-2207, 1991 31. Lord JT: Integration is key in consumer-centric approaches. Manag Care Interface 15:34-37, 2002 32. Humana looks to ISO registration to address quality improvement and customer satisfaction. Qual Lett Healthc Lead 15:8-9, 2003[Medline] 33. Joseph NE, Sigurdson ER, Hanlon AL, et al: Accuracy of determining nodal negativity in colorectal cancer on the basis of the number of nodes retrieved on resection. Ann Surg Oncol 10:213-218, 2003[CrossRef][Medline]
34. Wong JH, Severino R, Honnebier MB, et al: Number of nodes examined and staging accuracy in colorectal carcinoma. J Clin Oncol 17:2896-2900, 1999 35. Suzuka I, Shiota K, Yamane M, et al: Intra-operative detection of lymph node involvement in carcinoma of the colon. Hepatogastroenterology 45:2127-2131, 1998[Medline] 36. Fédération Nationale des Centres de Lutte Conter le Cancer: Cancers du Sein Non Métastatiques (vol 3, ed 1). Paris, France, Fédération Nationale des Centres de Lutte Conter le Cancer, 1996, pp 5-348 37. Mauriac L, Luporsi E, Cutuli B, et al: Summary version of the standards, options and recommendations for nonmetastatic breast cancer (updated January 2001). Br J Cancer 89:S17-S31, 2003 (suppl 1) 38. Delamothe T: Wanted: Guidelines that doctors will follow. BMJ 307:218, 1993 Submitted September 14, 2004; accepted March 8, 2005.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 or Fisher's exact tests, as appropriate. Continuous data were analyzed with the Student's t test. The statistical significance level was set at P = .05 in a two-sided test. Univariate analyses were performed to assess whether patient characteristics were correlated with compliance or noncompliance with the CPG. A Mantel-Haenszel 
.003) in the experimental group, whereas this was true only for the metastatic stage in the control group (P < .001) and not the case for localized stages (P = .11). 
