|
|
| |||||||||||||||||||||||||||||||||||||||
Background Case-control studies have demonstrated that screening by sigmoidoscopy is effective in reducing mortality from colorectal cancer. If nurses performed screening examinations, more patients could be screened and, at current income levels, at a lower cost.
Methods Two registered nurses and two licensed practical nurses learned to perform examinations with the flexible fiberoptic sigmoidoscope in order to screen patients for colorectal tumors. They performed 1881 independent examinations of outpatients more than 45 years of age. During the same period, 730 examinations were performed by two gastroenterologists in similar patients.
Results The mean depth of insertion of the sigmoidoscope was slightly but significantly greater in the patients examined by the physicians than in those examined by the nurses (48 vs. 46 cm in men, P = 0.003; 41 vs. 38 cm in women, P = 0.002). Adenomas were found in 14 percent of the men and 8 percent of the women examined (P = 0.001). Nine cancers were found in men and four in women. There were no significant differences between the nurses and the physicians in the proportion of examinations that were positive for adenomas or cancer. No complications occurred during the initial examinations or during 894 follow-up sigmoidoscopic procedures. Among the patients whose initial examination results were normal, more of those examined by nurses returned for follow-up sigmoidoscopy after 12 months or more (45 percent, vs. 30 percent of those examined by physicians; P = 0.001).
Conclusions Nurses can carry out screening by flexible sigmoidoscopy as accurately and safely as experienced gastroenterologists.
The American Cancer Society,4 the National Cancer Institute,5 and the American College of Physicians6 recommend that asymptomatic adults older than 50 years of age undergo screening by sigmoidoscopy to detect adenomas and early cancer. Conversely, the U.S. Preventive Services Task Force7 and the Canadian Task Force on the Periodic Health Examination8 do not support such screening because there have been no data from controlled studies that demonstrate a survival advantage in the screened group.
Impediments to widespread screening include inadequate numbers of trained practitioners and the expense of the examination. A considerable increase in the availability of screening and a reduction in cost could be achieved if nurses performed screening sigmoidoscopy instead of physicians, and if the difference in cost between nurses and physicians were maintained. My colleagues and I undertook this study to determine whether screening by flexible sigmoidoscopy could be performed safely and accurately by nurses.
Methods
Training Program for Nurses
Two registered nurses and two licensed practical nurses volunteered to learn to perform flexible sigmoidoscopy. First, each nurse read a standard textbook on the rationale for screening and the technique of sigmoidoscopy9 and reviewed a collection of 35-mm slides showing endoscopic anatomy. The nurse gained proficiency with the controls of a 60-cm fiberoptic flexible sigmoidoscope (Olympus CF-P10S, Olympus, Lake Success, N.Y.) during supervised practice with a plastic model of the rectum and sigmoid colon (Olympus). She then used a fiberoptic teaching attachment or a video monitor to observe the technique of a gastroenterologist who served as preceptor. Next, the nurse withdrew the instrument after it had been inserted by the preceptor. Finally, the nurse performed a complete examination. During withdrawals and complete examinations, the preceptor supervised each procedure, using the teaching attachment or video system. The completion of each stage of the training depended on the preceptor's judgment of the nurse's performance. Approximately 35 observation trials, 30 withdrawal trials, and 35 supervised complete examinations during a period of three to five weeks were required before independent examinations were attempted. All the nurses successfully completed training, and the results of all their independent examinations are included here.
Sigmoidoscopy was scheduled to be performed by a nurse only when the preceptor or another gastroenterologist was immediately available in an adjacent examination room. All suspicious lesions were observed by one of the gastroenterologists. Specimens from all raised lesions greater than 0.2 cm in diameter were obtained with standard biopsy forceps. The biopsy technique was taught when the nurses discovered lesions during independent examinations. The withdrawal portion of each independent examination performed by a nurse was recorded on videotape. For the first two years of the program, these videotapes were reviewed on a timely basis by the gastroenterologists.
Fiberoptic Flexible Sigmoidoscopy
Sodium phosphate enemas (C.B. Fleet, Lynchburg, Va.) were recommended two hours and one hour before the procedure. Written informed consent was obtained from each patient immediately before the examination. No anesthesia was used. The depth of insertion from the anal verge was estimated to the nearest 5 cm by comparison with a scale marked on the shaft of the instrument. Insertion to 60 cm was always attempted, but the procedure was stopped if stool compromised the examination or if the patient reported discomfort. The results of each procedure were entered into a computerized data base (The Endoscopy Database, MUMPS Medical Information Management Systems, Oak Park, Ill.) by the primary endoscopist immediately after the examination, and a computer-generated report was placed in the permanent medical record of each patient.
Patients
Screening by flexible sigmoidoscopy was encouraged for outpatients older than 45 years of age and was routinely performed by gastroenterologists at the Ochsner Clinic of Baton Rouge before the Nurse Clinic was established. Patients who had never undergone sigmoidoscopy at the clinic were eligible for referral to the Nurse Clinic. No attempt was made to determine whether the patients had previously undergone sigmoidoscopy elsewhere. We requested that each patient have six stool samples tested for occult blood before sigmoidoscopy, but many patients were referred without documentation of testing. We also suggested that patients with a history of colorectal cancer or adenomas and patients with symptoms related to the colon be referred to the Gastroenterology Clinic, but many patients with symptoms or bleeding were examined in the Nurse Clinic. Patients in whom cancer or adenomas were detected during sigmoidoscopy were offered outpatient colonoscopy. No attempt was made to remove tumors during sigmoidoscopy.
Assessment of Patients' Perceptions
Immediately after the first procedure, some patients were asked to complete forced-choice visual-analogue scales10 to measure their perceptions of the severity of cramps, burning, pressure, bloating, and embarrassment during sigmoidoscopy and their perceptions of the explanation of the procedure and the results before and after sigmoidoscopy. The patients were also asked to estimate when they would be willing to undergo reexamination. Six possible responses were allowed for each item in the scale. The first 128 patients examined in the Nurse Clinic and 64 consecutive patients examined in the Gastroenterology Clinic during the same six-week period served as the first sample. A second sample of 118 consecutive patients from the Nurse Clinic and 28 consecutive patients from the Gastroenterology Clinic was tested three months later.
Statistical Analysis
The relations between categorical variables were assessed with the Mann-Whitney U test, the chi-square test, or Kruskal-Wallis analysis of variance of the ranks. The relations between continuous variables were assessed with Student's t-test. Analyses were performed on GBSTAT statistical-analysis software (Dynamic Microsystems, Silver Spring, Md.). All reported P values are two-tailed.
Results
A total of 1475 women (mean [±SD] age, 60 ±10 years) and 1136 men (61 ±10 years) underwent their first sigmoidoscopy at the clinic as a part of this study. The ages of the 1881 patients examined by the nurses were similar to those of the 730 patients examined by the physicians. Fifty-six percent of the patients examined by the nurses and 57 percent of those examined by the physicians were women. The only complication was in a 51-year-old man who had abdominal pain 24 hours after being examined by a nurse endoscopist. He was admitted to the hospital, where his symptoms resolved without fever or leukocytosis. The results of serial abdominal radiographic examinations were normal.
The mean depth of insertion recorded by the two physicians was greater than the mean depth recorded by the four nurses in examinations of both women (41 vs. 38 cm, P = 0.002) and men (48 vs. 46 cm, P = 0.003). The endoscope was inserted to 60 cm twice as frequently in men as in women (Figure 1). In men, the first peak of the distribution was at 40 cm, the depth just proximal to the junction of the sigmoid colon and the descending colon. Successful passage of this area usually insured insertion to 60 cm, but when the angle between the sigmoid colon and the descending colon was acute, further insertion caused discomfort and the examination was frequently stopped. In women, the peak at 40 cm was blunted by difficulty passing the junction of the rectum and the sigmoid colon at 20 cm. The differences in depth of insertion between men and women were present in patients of all ages. More examinations in women than men were stopped because of discomfort (Table 1), but there were no significant differences in the proportion stopped because of the presence of stool.
|
|
0.4 cm) declined colonoscopy. Nine carcinomas were found in men and four in women (P = 0.20). One cancer was found during colonoscopy after a 0.6-cm tubular adenoma had been sampled during sigmoidoscopy. Among the study patients, 1249 women (85 percent) and 901 men (79 percent) with normal examinations underwent their first sigmoidoscopy at least 12 months before the end of the study; 519 of these women (42 percent) and 375 of these men (42 percent) returned for a follow-up examination after 13 to 53 months. More patients initially examined by nurses (45 percent) than patients examined by physicians (30 percent, P = 0.001) returned for follow-up. The mean depth of insertion during follow-up sigmoidoscopy was 41 cm for women and 49 cm for men. There were no significant differences in depth of insertion between physicians and nurses.
Videotape review revealed that a single, small tubular adenoma (<1 cm) was missed during the first examination. However, 16 women (3 percent) and 29 men (8 percent) had tumors at follow-up sigmoidoscopy that were not detected during the first sigmoidoscopic examination (Table 2). Forty-three were tubular adenomas (all but three of which were less than 1 cm in diameter), one 0.8-cm adenoma had villous architecture, and one 0.8-cm lesion contained carcinoma. The carcinoma, a Dukes' stage A lesion (T1N0M0), was located just inside the anal verge and was found 16 months after normal results on an examination performed by a nurse.
|
The discomfort scores were low, and most patients rated the explanations given before and after the examination as excellent. The majority of patients reported that they would be willing to undergo a repeat sigmoidoscopic examination within 6 to 12 months, if indicated. In the first sample, patients examined by a physician had more cramping than those examined by a nurse (P = 0.001) and had a greater mean depth of insertion (45 cm vs. 39 cm, P = 0.001). However, there was no correlation between cramping and depth of insertion. Patients whose examinations were stopped because of discomfort had higher cramping scores than those in whom insertion to 60 cm was achieved (P = 0.003). There were no other significant differences between physicians and nurses in any item on the rating scale. Among the patients in the second sample, the depth of insertion in the two groups was equivalent, and there were no differences between nurses and physicians in any item on the scale.
Discussion
No randomized clinical trial has demonstrated that screening by sigmoidoscopy improves survival in patients with colorectal cancer, but two recent reports reinforce the case for screening. In a study of 261 patients who died of cancer located within 20 cm of the anus and 868 control patients, the odds ratio of having undergone screening by sigmoidoscopy was 0.3 in the patients who died of cancer, as compared with the control patients (95 percent confidence interval, 0.2 to 0.5)11. In a smaller study,12 the odds ratio of having undergone screening by sigmoidoscopy was 0.2 among 66 patients who died of distal colorectal cancer, as compared with 196 control patients. If the U.S. Preventive Services Task Force7 were to reevaluate the quality of the data supporting screening by sigmoidoscopy today, we believe that periodic screening of patients at average risk for colorectal cancer would probably be recommended.
Two reports have evaluated the ability of nurses to perform screening by flexible sigmoidoscopy. In one report, a registered nurse practitioner who was trained to use a 60-cm instrument performed 825 examinations with no complications13. The mean depth of insertion was 50 cm. In another study, Schroy et al.14 trained a nurse practitioner to use a prototype 40-cm video sigmoidoscope; he then examined 100 patients with no complications. The mean depth of insertion was 38 cm.
In our study nurse endoscopists safely inserted flexible sigmoidoscopes to depths similar to those reported by colorectal surgeons,15,16 gastroenterologists,17 and primary care physicians in practice18 and in training19,20. We found significant differences in the mean depth of insertion between men and women patients that were greater than the differences between physician and nurse endoscopists. Meaningful differences in the depth of insertion between men and women during screening by flexible sigmoidoscopy have not been reported previously, but in one report the depth of insertion during symptom-limited rigid sigmoidoscopy was greater in men than in women21. The differences in the prevalence of adenomas between men and women were not explained by differences in depth of insertion, because 90 percent of the lesions were located within 40 cm of the anus. Others have also reported that distal adenomas and cancer occur more frequently among men22,23.
The sensitivity of screening by sigmoidoscopy has not been rigorously evaluated. When successive colonoscopic procedures were performed,24 12 percent of adenomas less than 1.0 cm in diameter and situated in the sigmoid or descending colon were missed by the first examiner. When patients with adenomas underwent colonoscopy within three months after diagnostic sigmoidoscopy, 5 percent of adenomas greater than 0.5 cm in diameter that were found in the rectum and sigmoid colon had been missed during sigmoidoscopy25. Among 894 of our patients who returned 13 to 53 months after a normal sigmoidoscopy, fewer than 5 percent had tumors in the distal colon. My colleagues and I do not know how many of the lesions we found at the second examination were present during the first. However, only four large tubular adenomas, one villous adenoma, and one invasive cancer were noted during the second examination; the remainder were small tubular adenomas ( 1 cm). The risk of cancer in small tubular adenomas is low,26 and the risk of proximal cancer in patients with small tubular adenomas is no higher than that in the general population23,27.
The patients in our study accepted sigmoidoscopy performed by nurses without complaint. However, the nurse endoscopists had considerably more supervised practice than is usually recommended for nonendoscopist physicians who are learning sigmoidoscopy28,29,30. We were surprised by the number of examinations terminated early because of the patient's discomfort, given the low pain scores recorded by patients immediately after the procedure. Apparently, in our relatively asymptomatic patients, even low levels of discomfort were not acceptable. It is not clear why sigmoidoscopy was terminated early because of pain in more women patients, but anatomical differences, trauma occasioned by past pregnancy, or adhesions from pelvic surgery21 probably contributed to procedure-related discomfort among the women.
Except for the cost of training, few new expenses accrue to a program in which nurses perform screening by sigmoidoscopy. Most of its expenses, including those for salary and fringe benefits, are considerably lower than those incurred when physicians perform screening examinations. Since there is no evidence that physicians are more effective as endoscopists than nurses, and since screening by nurses could result in substantial savings, consideration should be given to the widespread training of nurses in sigmoidoscopy for screening purposes.
I am indebted to Michael Ruth, M.D., Elizabeth M. Ducote, R.N., Melanie G. Nutter, R.N., Marsha Breland, L.P.N., and Candace Marque, L.P.N., for their clinical skills and participation in this study; to William F. Waters, Ph.D., for assistance with the design of the analogue scale and with data analysis; and to Robert S. Sandler, M.D., Michael C. Perry, M.D., B. Jay Brooks, M.D., James H. Butt, M.D., and Carl G. Kardinal, M.D., who provided critical advice during the preparation of the manuscript.
Source Information
Presented in part at the fall meeting of the British Society of Gastroenterology, Southampton, United Kingdom, September 26, 1990.
From the Section of Gastroenterology, Department of Medicine, Ochsner Clinic of Baton Rouge, 16777 Medical Center Dr., Baton Rouge, LA 70816, where reprint requests should be addressed to Dr. Maule.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Nurse Practitioners as Endoscopists
Lahad A., Levy-Lahad E., Bryant J., Record N. B., Record S. S., Maule W. F.
Extract |
Full Text
N Engl J Med 1994;
330:1534-1535, May 26, 1994.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
