Evidence Table 1 - Question 1a (Published Literature): What quality-of-care measures are available and what evidence is available for these measures to assess the quality of diagnosis of colorectal cancer, including appropriate use of colon imaging, endoscopic visualization, and biopsy?

StudyCharacteristics of Quality MeasureTesting of Quality MeasureComments/Quality Scoring
Baig, Myers, Turner, et al., 2003 Quality measure (QM): Percentage of patients with a positive fecal occult blood test (+FOBT) who under-went a complete diagnostic evaluation (colonoscopy or double contrast barium enema [DCBE]) Study population: General comments:
#33210 Basis of QM: Clinical practice guideline; not cited, but appears to be Multi-Society Task Force (Winawer et al., Gastroenterology 1997, updated 2003; maybe also U.S. Preventive Services Task Force (Pignone et al., Ann Intern Med ~2002)HMO patients; primary care practicesMeasure has face validity because follow up of +FOBT with colonoscopy (or DCBE) is supported by practice guidelines, which themselves are supported by large well-designed RCTs with (disease-specific) mortality as the end point.
Type of QM: N: 544Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 4
  • Scientific acceptability (five criteria):
    • Precise specifications: 4
    • Reliability: 4
    • Validity: 4
    • Adaptability: 4
    • Adequacy of risk adjustment: 1
(a) ProcessAge: ≥ 50
(b) GeneralRace: NR
Outcome to which the QM is linked: Mortality in prior trials - no link within this particular paperSex: NR
Intent of QM: Not specified (aim of the study was to determine reasons for non-performance of a complete diagnostic evaluation)Tumor stage: NA
Definition of denominator/numerator: Geographic location: Greater Philadelphia, PA
Denominator:Dates: FOBT result, Aug-Nov 1998; follow up at 60 days
Number of Aetna Health (an HMO in greater Philadelphia, PA) participants aged 50 and older who had a +FOBT from Aug-Nov 1998Healthcare setting: HMO primary care practices
Numerator: Results:
Number above subjects who had a colonoscopy or DCBE at 60 days from the +FOBT, as indicated by mailed survey and claims dataOf the 544 +FOBT patients, 46% did not receive colonoscopy or DCBE.
Data sources: Provider survey, MCO claims data
Recommended frequency of data collection: No recommendation
Beart, Steele Jr., Menck, et al., 1995 Quality measure (QM): Percentage of patients with colon (or rectal) cancer undergoing colonoscopy as part of their initial evaluation Study population: General comments:
#990 Basis of QM: NoneN: 39,502 (29,209 colon; 10,293 rectal)The data for this study were generated prior to 1990 so it may not reflect more modern recommendations for management.
Type of QM: ColonRectumThis study reports the frequency of use for a number of diagnostic tests (chest x-ray, colonoscopy, biopsy of primary site, barium enema, preoperative CEA, LFTs, CT scan of primary site or liver, IVP, chest CT, etc.) that were performed for patients with colon and rectal cancer, but it is never stated which ones should have been considered components of quality care. Therefore, it is not possible to determine which are quality measures, and none of these are reported in this abstraction except colonoscopy, which seems to have face validity as a quality measure.
(a) Process 1983198819831988Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 4
  • Usability: 5
  • Scientific acceptability (five criteria):
    • Precise specifications: 2
    • Reliability: 4
    • Validity: 4
    • Adaptability: 4
    • Adequacy of risk adjustment: 2
(b) GeneralCases12,68216,5274,5975,696
Outcome to which the QM is linked: None is specified, but presumably it would be the finding of synchronous lesions that should be dealt with during surgery as well as the primary tumor< 50 yrs5.65.36.16.1
Intent of QM: Not specified (aim of the study was to identify current trends in the management of patients with carcinoma of the colon or rectum and to identify changes in patterns of care and survival)50–6939.437.446.146.0
Definition of denominator/numerator: 70–7933.433.931.531.2
Denominator:80+210623.416.216.7
Number of cases of colon and rectal cancer (up to 25 per program or facility) reported by over 1,200 approved cancer programs and 800 other facilities on the Commission on Cancer mailing list that were invited to participate.Male %47.849.355.957.2
Numerator:Tumor Stage
Number of above cases having colonoscopy as part of their initial evaluation.03.75.55.66.1
Data sources: Hospitals sent in up to 25 consecutive cases from their medical recordsI21.222.927.428.3
Recommended frequency of data collection: No recommendationII29.928.622.721.6
III22.422.121.922.0
IV20.117.717.016.7
Unkown2.73.35.45.4
Non-hispanic white 86.1 87.0
Hispanic 2.7 3.5
AA 9.1 6.5
Asian 1.1 1.6
Performance status: NR
Geographic location: Entire U.S.
Dates: Calendar years 1983 and 1988
Healthcare setting: Hospitals
Results:
44% of colon cancer patients in 1983 and 63.8 in 1988 underwent colonoscopy.
30.1% of rectal cancer patients in 1983 and 54.6% in 1988 underwent colonoscopy.
Harewood, Sharma, and de Garmo, 2003 Quality measure (QM): Percentage of patients with adequate preparation of bowel prior to colonoscopy Study population: General comments:
#33050 Basis of QM: NoneN: 93,004The ability to achieve adequate colonoscopic preparation is predominantly patient-dependent.
Type of QM: Age: NRThis paper also does not exclusively deal with colon cancer and does not discuss the reasons the colonoscopy was performed.
(a) ProcessRace: NRRating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 2
  • Usability: 5
  • Scientific acceptability (five criteria):
    • Precise specifications: 4
    • Reliability: 3
    • Validity: 3
    • Adaptability: 3
    • Adequacy of risk adjustment: NA
(b) TechnicalSex: NR
Outcome to which the QM is linked: Detection of colonic lesionsTumor stage: NR; most cases were not colorectal cancer
Intent of QM: Not specified (aim of the study was to characterize the impact of bowel preparation adequacy on detection of colonic lesions)Performance status: NR
Definition of denominator/numerator: Geographic location: 88 sites in 24 states
Denominator:Dates: Jan 1, 2000 to Dec 31, 2001
Number of colonoscopies performed from January 1, 2000, to December 31, 2001, that were completed and had complete documentation of age, sex, preparation quality, and endoscopic findings.Healthcare setting: Varied (community, academic, and VA institutions)
Numerator: Results:
Number of above colonoscopies for which bowel preparation was rated as adequate.76.9% of colonoscopies with complete data were rated as having an adequate preparation.
Data sources: Clinical Outcomes Research Initiative (CORI) database, which collects data from endoscopic procedures from a consortium of 580 specialists in GI diseases at 88 sites in 24 statesRegarding the relationship of adequacy of bowel preparation to ability to detect colonic polyps/cancers: Suspected neoplasia was detected in 29.1% of colonoscopies with adequate preparation and 26.4% with inadequate preparation (P < 0.0001).
Recommended frequency of data collection: No recommendationPolyps > 9 mm were identified with equal frequency (7.3%) regardless of adequacy of preparation. Polyps ≤ 9mm were identified in 21.8% of colonoscopies with adequate preparation and 19.0% with inadequate preparation (P < 0.0001).
After adjusting for age and sex, adequate preparation was predictive of detection of suspected neoplasia.
Haseman, Lemmel, Rahmani, et al., 1997 Quality measure (QM): Percentage of colon cancers not detected by colonoscopy in symptomatic patients. (Only 7% of the patients in this study had routine screening colonoscopies.) Study population: General comments:
#35910 Basis of QM: NoneN: 941Data for this study were collected from hospitals that “agreed to participate,” which raises concern about whether bias may occur. Only hospitals with low “miss” rates might want to participate; others might not to avoid “bad publicity.”
Type of QM: Age: 69.8Cecal intubation rates were not calculated, but would be relevant because failure to reach the cancer is a factor in failed detection.
(a) ProcessRace: NRData on size of tumors were not collected.
(b) TechnicalSex: NRRating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 5
  • Scientific acceptability (five criteria):
    • Precise specifications: 4
    • Reliability: 4
    • Validity: 3
    • Adaptability: 4
    • Adequacy of risk adjustment: 3
Outcome to which the QM is linked: Not stated but presumably linked to stage of colon cancer that a patient is ultimately diagnosed withTumor stage: Various
Intent of QM: Not specified, but presumably to help with future quality improvement projects (aim of the study was to understand the reasons for colonoscopic failures to detect colon cancer and polyps)Performance status: NR
Definition of denominator/numerator: Geographic location: Hospitals in central Indiana
Denominator:Dates: 1988-1993
Number of cases of colorectal cancer in which colonoscopy was used within 3 years of diagnosis.Healthcare setting: Hospitals
Numerator: Results:
Number of above cases of colon cancer that were missed by the colonoscopy.Failure to detect colorectal cancer by colonoscopy occurred in 5%. All undetected tumors were “missed” rather than not reached.
Data sources: Data were collected from 20 hospitals that agreed to participate in this project
Recommended frequency of data collection: No recommendation
Levin, Hess, and Johnson, 1997 Quality measure (QM): Number of participants undergoing adequate diagnostic follow up of positive fecal occult blood test (+FOBT) Study population: General comments:
#11880 Basis of QM: Clinical practice guideline (ACS guidelines [ref 16])N: 934The purpose of this paper was to compare three different hemoccult tests for detecting fecal occult blood, but it also obtained information on whether diagnostic followup was obtained.
Type of QM: Age: 50–94Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 5
  • Scientific acceptability (five criteria):
    • Precise specifications: 4
    • Reliability: 3
    • Validity: 3
    • Adaptability: 3
    • Adequacy of risk adjustment: 1
(a) ProcessRace: 90% white, 4.8% black, 2.9% Hispanic, 1.4% Asian, < 1% American Indian
(b) GeneralSex: 43.3% male, 56.7% female
Outcome to which the QM is linked: Detection of colonic neoplasmsTumor stage: NR
Intent of QM: Not specified (aim of the study was to compare three FOBTs and assess patient and physician compliance with ACS guidelines on recommended diagnostic workup of participants with a +FOBT)Performance status: NR
Definition of denominator/numerator: Geographic location: Houston, TX
Denominator:Dates: 4-week period in October 1993
Number of participants with +FOBT tests for whom followup information was available.Healthcare setting: Not stated; all were outpatients when participating in the study
Numerator: Results:
Number of above participants undergoing adequate diagnostic follow up (colonoscopy or double-contrast barium enema with flexible sigmoidoscopy)Rate of adequate diagnostic followup was 59% overall and varied depending on specialty of the physician who was consulted:
Data sources: Telephone calls to participants; questionnaires sent to physiciansFamily physician: 45%
Recommended frequency of data collection: No recommendationOther physician: 55%
Gastroenterologist: 85%
Neoplasia was discovered in 56/553 (10%) participants who received adequate diagnostic follow up regardless of physician type, but in only 4/553 (1%) participants who did not get adequate followup care. This difference was statistically significant (p < 0.001).
Mandel, Church, Bond, et al., 2000 Quality measure (QM): Percentage of patients who have a followup evaluation of a positive fecal occult blood test (+FOBT) Study population: General comments:
#16070 Basis of QM: Clinical practice guideline Annual ScreeningBiennial ScreeningThe purpose of this paper was to present data on the cumulative incidence of colorectal cancer in the individual groups who were screened by annual, biennial, or usual care. It does report the percentage of patients who had follow up of a positive FOBT.
Type of QM: Male7,4747,430Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance:
  • Usability: 5
  • Scientific acceptability (five criteria):
    • Precise specifications: 5
    • Reliability: 5
    • Validity: 5
    • Adaptability: 3
    • Adequacy of risk adjustment: 1
(a) ProcessFemale8,0588,120
(b) GeneralTotal15,53215,550
Outcome to which the QM is linked: Positive diagnosis of colorectal cancerAverage age: 50–80 years old
Intent of QM: Quality improvement, researchGeographic location: Minnesota
Definition of denominator/numerator: Dates: 1976-1992
Denominator:Healthcare setting: Randomized trial
Number of patients with a +FOBT. Results:
Numerator:83% of patients in the annual-screening group and 84% of patients in the biennial-screening group underwent diagnostic followup, including a complete examination of the large bowel by colonoscopy or the combination of double-contrast enema and flexible sigmoidoscopy. In each group about 11% of the subjects with positive screening tests underwent flexible sigmoidoscopy or barium enema or underwent another FOBT. Five percent of the subjects with positive tests declined to consult a physician.
Number of above patients who had appropriate diagnostic evaluation.
Data sources: Database of the Minnesota Colon Cancer Control study
Recommended frequency of data collection: No recommendation
Marble, Banerjee, and Greenwald, 1992 Quality measure (QM): Time from patient presentation to physician diagnosis Study population: General comments:
#4910 Basis of QM: NoneN: 50/50A small study of limited use in quality measurement.
Type of QM: Age: ≤ 40/ > 40Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 3
  • Usability: 4
  • Scientific acceptability (five criteria):
    • Precise specifications: 2
    • Reliability: 1
    • Validity: 1
    • Adaptability: 3
    • Adequacy of risk adjustment: 1
(a) ProcessRace: 94% white/98% white
(b) GeneralSex: 58% female/52% female
Outcome to which the QM is linked: More rapid diagnosis presumed to be linked to earlier stage at diagnosis and hence improved cure rates. (Unlikely that there are data to support this idea over the typical range of time between presentation and diagnosis.)Tumor stage: NR
Intent of QM: Not specified (aim of the study was to determine whether younger patients with colorectal cancer had a poorer prognosis than their older counterparts, and if they did, the reasons underlying this)Performance status: NR
Definition of denominator/numerator: Geographic location: Hartford, CT
Denominator:Dates: 1935-1988
Numerator:Healthcare setting: Hospital
Data sources: Saint Francis Hospital and Medical Center Tumor Registry Results:
Recommended frequency of data collection: No recommendationThere was no difference in the interval from presentation to diagnosis in older vs. younger patients (the appropriate diagnosis was made in < 1 week in > 90% of the patients in both groups).
Myers, Turner, Weinberg, et al., 2004 Quality measure (QM): Percentage of patients receiving complete diagnostic evaluation (CDE) for positive fecal occult blood test (+FOBT), defined as colonoscopy or barium enema + flexible sigmoidoscopy (BEFS) Study population: General comments: None
#30300 Basis of QM: Clinical practice guideline (Winawer et al., Gastroenterology, 2003)N: 2,992Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 4
  • Scientific acceptability (five criteria):
    • Precise specifications: 4
    • Reliability: 4
    • Validity: 4
    • Adaptability: 4
    • Adequacy of risk adjustment: 1
Type of QM: Age: ≥ 50
(a) ProcessGeographic location: Southern PA and NJ
(b) GeneralDates: 1994-2000, 3 study periods
Outcome to which the QM is linked: Reduced colorectal cancer-related mortalityHealthcare setting: Aetna U.S. Healthcare and various MCO-based cancer screening sites
Intent of QM: Not specified (aim of the study was to evaluate the impact of a guideline on CDE performance rates in primary care practices) Results:
Definition of denominator/numerator: Intervention groupControl group
Denominator:CDE recommendation rate79.667.3
Number of +FOBT patients eligible for CDE recommendation and performance. Exclusion criteria: CDE procedures completed within 3 years prior to +FOBT result; patient had a medical condition that contraindicated CDE; patient deceased; patient unknown to practice and patient left practice before CDE could be recommended.CDE performance rate63.353.7
Numerator:For study period 3 (5/99 to 2/00), the differences in CDE recommendation and performance rates between the intervention group and control group was statistically significant. Use of a physician-oriented intervention substantially and significantly increased CDE recommendation and performance in intervention group practice as compared to control group practices. Targeting PCPs for delivery of a combined CDE reminder feedback and educational outreach intervention can have a meaningful impact on physician behavior and patient followup in colorectal cancer screening.
Number of above patients whose physicians had recommended CDE; number of patients who underwent CDE within 180 days after the +FOBT result. Results were coded as dichotomous variables, yes or no (recommendation was made or treatment was provided).
Data sources: ICA forms; MCO administrative claims data
Recommended frequency of data collection: No recommendation
Nelson, McQuaid, Bond, et al., 2002 Quality measure (QM): Postprocedural complication rateStudy population: n = 3,196General comments: None
#45970 Basis of QM: Not specifiedMean age63.0Rating of quality measure as presented in this study(scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 3
  • Scientific acceptability (five criteria):
    • Precise specifications: 5
    • Reliability: 4
    • Validity: 4
    • Adaptability: 4
    • Adequacy of risk adjustment: 4
Type of QM: % male96.8
(a) Outcome% white83.5
(b) TechnicalComorbidities:
Intent of QM: Not specified (aim of the study was to report and identify predictive variable for procedural success and complication rates of screening colonoscopy in a large asymptomatic cohort)Coronary heart disease21.1%
Definition of denominator/numerator: CVA/TIA8.2%
Denominator:Diabetes20.8%
Number of asymptomatic patients between 50 and 75 years undergoing a screening colonoscopy at 13 VAMCs between 2/1994-1/1997.COPD8.5%
Exclusion criteria: Patients were excluded if they reported symptoms of lower GI tract disease, including rectal bleeding on more than one occasion in the prior 6 months, significant change in bowel habits or lower abdominal pain that would require evaluation. Also excluded were those with prior colonic disease; prior colon examination within 10 years; significant medical problems that would increase risk of colonoscopy, or a medical condition that would preclude benefit from screening; need for special precautions for colonoscopy; women with childbearing potential.Geographic location: 13 VAMCs across U.S.
Numerator:Dates: 1994-1997
Number of above cases where postprocedural complications within 30 days occurred, including perforation, GI bleeding with hospitalization, new arrhythmia, MI/CVA, death within 30 days.Healthcare setting: General medicine clinics
Data sources: Medical chart Results:
Recommended frequency of data collection: No recommendationNo statistically significant complications occurred.
Nelson, McQuaid, Bond, et al., 2002 Quality measure (QM): Percent of patients with “adequate” or better bowel preparationStudy population: n = 3,196General comments: None
#45970 Basis of QM: Not specifiedMean age63.0Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 3
  • Scientific acceptability (five criteria):
    • Precise specifications: 5
    • Reliability: 4
    • Validity: 4
    • Adaptability: 4
    • Adequacy of risk adjustment: 4
Type of QM: % male96.8
(a) Outcome% white83.5
(b) TechnicalComorbidities:
Outcome to which the QM is linked: Coronary heart disease21.1%
Intent of QM: Not specified (aim of the study was to report and identify predictive variable for procedural success and complication rates of screening colonoscopy in a large asymptomatic cohort)CVA/TIA8.2%
Definition of denominator/numerator: Diabetes20.8%
Denominator:COPD8.5%
Number of asymptomatic patients between 50 and 75 years undergoing a screening colonoscopy at 13 VAMCs between 2/1994-1/1997.Geographic location: 13 VAMCs across U.S.
Exclusion criteria: Patients were excluded if they reported symptoms of lower GI tract disease, including rectal bleeding on more than one occasion in the prior 6 months, significant change in bowel habits or lower abdominal pain that would require evaluation. Also excluded were those with prior colonic disease; prior colon examination within 10 years; significant medical problems that would increase risk of colonoscopy, or a medical condition that would preclude benefit from screening; need for special precautions for colonoscopy; women with childbearing potential.Dates: 1994-1997
Numerator:Healthcare setting: General medicine clinics
Number of above cases where bowel preparation was rated by the endoscopist as “good” (mucosa well seen throughout), “fair” (liquid, contents; exam adequate); or ”poor” (solid contents, exam compromised). Results:
Data sources: Medical chartBy using a polyethylene glycol-based electrolyte solution, the bowel preparation was described as good in 81.4%, fair in 15.8% and poor in 2.7% of patients.
Recommended frequency of data collection: No recommendation.
Nelson, McQuaid, Bond, et al., 2002 Quality measure (QM): Percent of patients with a colonoscopy with successful cecal intubationStudy population: n = 3,196General comments: None
#45970 Basis of QM: Not specifiedMean age63.0Rating of quality measure as presented in this study(scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 3
  • Scientific acceptability (five criteria):
    • Precise specifications: 5
    • Reliability: 4
    • Validity: 4
    • Adaptability:
    • Adequacy of risk adjustment: 4
Type of QM: % male96.8
(a) Process% white83.5
(b) TechnicalComorbidities:
Outcome to which the QM is linked: Coronary heart disease21.1%
Intent of QM: Not specified (aim of the study was to report and identify predictive variable for procedural success and complication rates of screening colonoscopy in a large asymptomatic cohort)CVA/TIA8.2%
Definition of denominator/numerator: Diabetes20.8%
Denominator:COPD8.5%
Number of asymptomatic patients between 50 and 75 years undergoing a screening colonoscopy at 13 VAMCs between 2/1994-1/1997.Geographic location: 13 VAMCs across U.S.
Exclusion criteria: Patients were excluded if they reported symptoms of lower GI tract disease, including rectal bleeding on more than one occasion in the prior 6 months, significant change in bowel habits or lower abdominal pain that would require evaluation. Also excluded were those with prior colonic disease; prior colon examination within 10 years; significant medical problems that would increase risk of colonoscopy, or a medical condition that would preclude benefit from screening; need for special precautions for colonoscopy; women with childbearing potential.Dates: 1994-1997
Numerator:Healthcare setting: General medicine clinics
Number of above cases where colonoscopy with procedure was successful; that is, confirmation of cecal intubation. Results:
Data sources: Medical chartColonoscopy with cecal intubation was successful on the first attempt in 97.2% of the cases. This includes 69 cases in which the quality of the preparation was felt to be inadequate to visualize the entire colonic mucosa. 53.8% of these cases had at lest 1 polyp resected.
Recommended frequency of data collection: No recommendation.
Shehadeh, Rebala, Kumar, et al., 2002 Quality measure (QM): Miss rate for followup colonoscopy after polypectomy Study population: General comments:
#18520 Basis of QM: NoneN: 122Because fellows in training performed most of the colonoscopies (albeit with gastroenterologist supervision), the results may not be reflective of routine clinical practice.
Type of QM: Age: NRMiss rates could be used for internal quality improvement purposes (as was done in this case).
(a) ProcessRace: NRRating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 5
  • Scientific acceptability (five criteria):
    • Precise specifications: 5
    • Reliability: 3
    • Validity: 5
    • Adaptability: 3
    • Adequacy of risk adjustment: 1
(b) TechnicalSex: 100% male
Outcome to which the QM is linked: Presumably, improved survival via earlier detection of malignancyTumor stage: NA
Intent of QM: Not specified (aim of the study was to evalute the miss rates of advanced adenomas)Performance status: NR
Definition of denominator/numerator: Geographic location: Dayton, Ohio
Denominator:Dates: 1992-1999
Number of patients who had polypectomy on initial colonoscopy and who had at least one followup colonoscopy from July 1, 1992 to June 30, 1999 at the Dayton VAMC. Colonoscopy had to be complete and bowel prep had to be satisfactory.Healthcare setting: Veteran's Affairs Medical Center (VAMC)
Numerator: Results:
Number of patients with missed advanced adenomas (≥10 mm) found on repeat colonoscopy.122 patients had 338 colonoscopies. Miss rates were calculated as in the National Polyp Study. Missed adenomas were defined as adenomas (especially big ones) found on repeat colonoscopy. 122 patients had 2 colonoscopies and 60 patients had a third colonoscopy.
Data sources: Computerized database and paper charts at VAMCThe calculated miss rate of advanced adenomas for the second colonoscopy was 2.5% (4/122); for the third colonoscopy, it was 3.3 %(2/60).
Recommended frequency of data collection: No recommendation
Shields, Weiner, Henry, et al., 2001 Quality measure (QM): Percentage of patients aged 40 or older participating in a mass screening program who had an “adequate evaluation” (colonoscopy or barium enema + flexible sigmoidoscopy [BEFS]) for a positive fecal occult blood test (+FOBT; Hemeoccult II)Study population: N: 940General comments: None
#23040 Basis of QM: Clinical practice guideline (American Cancer Society [ref 14, 15 1980 CA, 1996 JAMA]) percentRating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 4
  • Scientific acceptability (five criteria):
    • Precise specifications: 4
    • Reliability: 3
    • Validity: 4
    • Adaptability: 3
    • Adequacy of risk adjustment: 1
Type of QM: 40–49 yrs11
(a) Process50–59 yrs18
(b) General60–69 yrs37
Outcome to which the QM is linked: Reduced mortality from colorectal cancer70–79 yrs28
Intent of QM: Not specified (aim of the study was to determine whether factors like age, sex, and family history [among others] influence adequate evaluation of a +FOBT)80+ yrs6
Definition of denominator/numerator: male48
Denominator:female47
Number of patients who participated in a 1986 mass screening program with FOBT (pharmacy-based, processed at Beth Israel Hospital (Boston MA), who returned their cards, had at least 1 positive window (out of 6) and who provided (or their physician provided) followup data (90%).unspecified5
Numerator:Race: NR
Number of the above subjects who in addition had an “adequate evaluation” (colonoscopy or barium enema + flexible sigmoidoscopy [BEFS])Geographic location: Boston area, MA
Data sources: Physician survey “checklist,” patient or physician phone call (if survey/checklist not returned), pathology reports, patient survey at 3 yearsDates: 1986-1988
Recommended frequency of data collection: No recommendationHealthcare setting: Variety
Results:
59% of patients with a +FOBT were adequately evaluated. 11.2% of adequately evaluated patients had new colon cancers discovered.
The effects of age and sex on the adequacy of follow up were considered, but no significant differences were found by age group or between men and women.
Ure, Dehghan, Vernava 3 rd , et al., 1995 Quality measure (QM): Complication rate of colonoscopy in the elderly and non-elderly Study population: General comments:
#960 Basis of QM: NoneEldlerly/Non-elderlyThis paper reports the rate of the following quality measures of the performance of colonoscopy: colonoscopy completion, rate of perforations, and postpolypectomy bleeding, but some patients were undergoing colonoscopy for screening (16%).
Type of QM: N: 354/302Rating of quality measure as presented in this study (scale of 1–5, where 1 = poor, 3 = moderate, 5 = ideal):
  • Importance: 5
  • Usability: 5
  • Scientific acceptability (five criteria):
    • Precise specifications: 5
    • Reliability: 5
    • Validity: 5
    • Adaptability: 3
    • Adequacy of risk adjustment: 1
(a) OutcomeAge: ≥ 70 / 50–70
(b) TechnicalRace: NR
Intent of QM: Not specified (aim of the study was to evaluate the utility, morbidity, and patient tolerance of colonoscopy in elderly patients as compared to a similar group of younger patients)Sex: NR
Definition of denominator/numerator: Tumor stage: NA
Denominator:Performance status: NR
Number of elderly/non-elderly patients undergoing colonoscopy.Geographic location: St. Louis area
Numerator:Dates: “A recent 48-month period”
Number of the above patients with resulting complications.Healthcare setting: University Hospital
Data sources: Not specified Results:
Recommended frequency of data collection: No recommendationOverall morbidity was similar in elderly and non-elderly patients (24% vs. 16%, p = NS). Elderly patients were significantly more likely to require termination of the procedure because of inadequate bowel preparation or pain (17% vs. 1%, p < 0.05).
Colonoscopy was successful to the cecum or ileocolic anastomosis in 85% (555/656). No patient had a perforation; 2% postprocedure hemorrhage in those undergoing polypectomy.

From: Appendix E: Evidence Tables

Cover of Cancer Care Quality Measures
Cancer Care Quality Measures: Diagnosis and Treatment of Colorectal Cancer.
Evidence Reports/Technology Assessments, No. 138.
Patwardhan MB, Samsa GP, McCrory DC, et al.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.