Diane L. Manninen

Complications of Colonoscopy in an Integrated Health Care Delivery System

Context Data on the frequency of colonoscopy complications from population-based samples are lacking. Contribution The authors searched electronic health records at Kaiser-Permanente of Northern California for patients who died or who had complications due to colonoscopy within 30 days of the procedure. Almost all procedures were diagnostic or for surveillance of previous abnormal findings. Of16318 eligible procedures, 82 involved serious complications (5 in 1000 procedures). Of the 82 complications, 95% followed biopsy or removal of polyps, and 62% of the polyps removed were smaller than 10 mm. The perforation rate was 1 in 1000 procedures. One death was related to colonoscopy. Cautions Less than 1% of procedures studied were screening colonoscopies, so these complication rates might not apply to screening examinations. The Editors Colonoscopy is the final step in colorectal cancer screening, regardless of the initial test chosen, and is recommended for primary colorectal cancer screening in average- risk persons (14). Colorectal cancer screening targets apparently healthy people; therefore, the magnitude of the risk and severity of the harms from screening are important issues to consider when selecting a screening strategy (5). Described complications of colonoscopy include colonic perforation, postbiopsy and postpolypectomy bleeding, and postpolypectomy syndrome (a transmural colonic burn, marked by localized abdominal pain without evidence of frank perforation) (6). Diverticulitis, which is caused by a microscopic perforation of the colon, can also theoretically be caused by colonoscopy in persons with preexisting diverticulosis. Most estimates of colonoscopy complications come from referral centers (712) or closely monitored clinical trials (13), limiting the generalizability of the results to community practice. In a large series by a group of ambulatory endoscopy centers (14), endoscopists self-reported complications, possibly underestimating them (15). In this study, researchers were unable to evaluate postpolypectomy bleeding. Postpolypectomy bleeding is particularly difficult to assess in studies because its occurrence is often delayed. The U.S. Preventive Services Task Force (16), in a recent evidence review of colonoscopy complications, concluded that postpolypectomy bleeding was reported in relatively few studies and delayed bleeding was not reported at all. Studies using administrative databases typically lack access to detailed records, including indications, depth of insertion, and whether or how polyps are removed (17). For the present study, we relied on the automated data of Kaiser Permanente, Northern California (KPNC), an integrated health care delivery system. Colonoscopy was most often used to follow-up other tests, such as fecal occult blood tests, flexible sigmoidoscopy, or barium enema, or to conduct surveillance in persons with a personal or family history of colorectal cancer or colorectal adenoma. Few colonoscopies were performed for primary screening. We identified patients undergoing colonoscopy and followed them for 30 days after the procedure for hospitalization for procedure-related complications. For this analysis, we defined any procedure-related complication that led to hospitalization as a serious complication. Methods We used KPNC electronic medical records to select patients who had undergone colonoscopies between 1 January 1994 and 16 July 2002. This was an observational study, conducted in medical centers throughout the KP health care system, evaluating practice patterns as they existed at the time the included colonoscopies were performed by the endoscopists in the study. Electronic records were reviewed to identify immediate complications, outpatient visits, or hospital admission within 30 days of colonoscopy. Colonoscopies were included in the analysis if they were performed for patients 40 years of age or older who were undergoing coloscopy because of a family history of colorectal cancer or adenomatous polyp, as a follow-up to a positive screening test (that is, polyp or cancer at sigmoidoscopy, positive results on a fecal occult blood test, or abnormal barium enema radiography), for surveillance because of a previously detected adenomatous polyp or colorectal cancer, or for primary screening. Colonoscopies were not included if the procedure was being performed to diagnose symptoms (for example, diarrhea, abdominal pain, gastrointestinal bleeding, history of rectal bleeding, or anemia) or if patients had outpatient visits 6 months before the procedure for abdominal pain, anemia, diarrhea, or constipation. A total of 35945 procedures performed at KPNC between 1 January 1994 and 16 July 2002 were identified by using 2 electronic KPNC endoscopy databases. The first database, the Colorectal Cancer Prevention (CoCaP) program database, contains detailed information from 1994 to 1996 on sigmoidoscopies, follow-up colonoscopies, and the results of pathologic testing. Information available in the CoCaP database includes depth of insertion; size, number, and treatment of polyps; limitations of the procedure; and identity of the examiner. The second database, the EndoLog Pro database, includes colonoscopy reports from 1995 to 2002 from 5 KP facilities. The database contains information on number of polyps found and their treatment, depth of insertion, quality of bowel preparation, identity of the examiner, and any immediate complications. Some patients underwent more than 1 colonoscopy during the study period. If a colonoscopy was incomplete because of poor bowel preparation, and a second colonoscopy was performed within 3 months, only the second colonoscopy was included in the cohort. If a patient required a second colonoscopy to complete removal of a polyp, only the first colonoscopy was included in the cohort. Patients requiring frequent surveillance may have been screened more than once during the 7-year study period; colonoscopies were included for these patients if the interval between the colonoscopies was greater than 6 months. Identification of Eligible Cases Of the 35945 procedures, 4646 were excluded because patients were younger than 40 years of age; 9499 were excluded because the procedures were performed for excluded indications or for symptoms; and 2411 were excluded because of poor preparation (with a second examination rescheduled in 90 days), interval since previous procedure was less than 6 months, previous colon surgery, or because the procedure was for follow-up removal of residual polyps or for marking polyp site for surgery. Inpatient procedures (n= 125) and procedures for KPNC nonmembers (n= 57) were excluded. Procedures were also excluded if patients had inpatient or outpatient visits for lower gastrointestinal bleeding, abdominal pain, anemia, diarrhea, or constipation 6 months before the procedures (n= 2689). A total of 16318 procedures were included in the analysis. Identification of Possible Complications A 2-step procedure was used to identify serious complications. First, we analyzed KP electronic databases for evidence of patients being admitted to the hospital (a KPNC or nonprogram hospital) within 30 days of colonoscopy. We focused on admissions that could be associated with colonoscopy complications or complications of procedural sedation, including colonic perforation (International Classifications of Diseases, 9th revision, [ICD-9] codes 569.83 and 998.2]; lower gastrointestinal bleeding (ICD-9 558.9, 578.1, 995.2, 995.89, and 998.1 to 998.13); anemia, not explained by preexisting conditions (ICD-9 280.0 and 285.0 to 285.9); diverticulitis (ICD-9 562.11); colitis, not present during initial endoscopy (ICD-9 556 to 556.9); aspiration pneumonia (ICD-9 507); pneumonia, organism unspecified (ICD-9 486); infection (ICD-9 780.6, 790.7, and 424.9 to 424.99); abdominal pain (ICD-9 789.0 to 789.09); complications of procedure (E872, E872.8, E872.9, E879, E879.8, and E879.9); complications secondary to anesthesia (ICD-9 995.4, 997.1, and 997.3); myocardial infarction (ICD-9 410 to 410.92 and 414); and stroke (ICD-9 436). Deaths within 30 days of colonoscopy were identified through linkage with the National Death Index. After possible cases were identified from electronic records, medical records analysts at KPNC reviewed the hardcopy medical records, computerized medical records, and laboratory records of 183 patients by using chart review forms. Analysts made photocopies of histories and physicals, discharge summaries, colonoscopy reports, operative notes, and pathology reports, and these were used to make decisions. Two physicians reviewed the photocopied records to determine whether the hospitalization or death was related to colonoscopy. Clinical judgment was used in making these decisions through a collaborative process, and decisions were made by mutual agreement. A third physician adjudicated the 1 case in which there were ongoing questions. A subsample of 44 records was reviewed by both physicians independently. The -statistic for this statistical analysis was 0.71 (CI, 0.52 to 0.89). Statistical Analysis Individual complication measures were created to reflect the incidence of serious complications in the first 30 days after colonoscopy for the following: 1) colonic perforation; 2) the postpolypectomy syndrome; 3) bleeding requiring overnight hospitalization, overall and separately for patients with or without surgery or transfusion; 4) diverticulitis requiring overnight hospitalization, overall and separately for patients with or without surgery; and 5) any other hospitalization within 30 days that was likely to have been caused or exacerbated by the procedure. Two aggregate measures were used. The first was for all of the above categories combined and the second for the most serious complications, including perforation, bleeding with transfusion, and diverticulitis requiring surgery. For each complication measure, we calculated the inc

Model of Complications of NIDDM: II. Analysis of the health benefits and cost-effectiveness of treating NIDDM with the goal of normoglycemia

OBJECTIVE To analyze the health benefits and economics of treating NIDDM with the goal of normoglycemia. RESEARCH DESIGN AND METHODS Incidence-based simulation model of NIDDM was used. Hazard rates for complications were adjusted for glycemia using risk gradients from the Diabetes Control and Complications Trial. Treatment costs were estimated from national survey data and clinical trials. Incremental costs and benefits were expressed in present value dollars (3% discount rate). Life-years were adjusted for quality of life, yielding quality-adjusted life-years (QALYs). RESULTS Comprehensive treatment of NIDDM that maintains an HbA1c value of 7.2% is predicted to reduce the cumulative incidence of blindness, end-stage renal disease, and lower-extremity amputation by 72, 87, and 67%, respectively. Cardiovascular disease risk increased by 3% (no effect of treating glycemia is assumed). Life expectancy increased 1.39 years. The cost of treating hyperglycemia increased by almost twofold, which is partially offset by reductions in the cost of complications. The estimated incremental cost/QALY gained is

Use of Services by Diabetes Patients in Managed Care Organizations: Development of a diabetes surveillance system

16,002. Treatment is more cost-effective for those with longer glycemic exposure (earlier onset of diabetes), minorities, and those with higher HbA1c under standard care. CONCLUSIONS The incremental effectiveness of treating NIDDM with the goal of normoglycemia is estimated to be ∼

Do Adolescents Who Relinquish Their Children Fare Better or Worse Than Those Who Raise Them

16,000/QALY gained, which is in the range of interventions that are generally considered cost-effective.

Immunosuppressive therapy as a determinant of transplantation outcomes

OBJECTIVE Computerized inpatient, pharmacy, outpatient, and laboratory records were used to develop an algorithm to identifydiabetes patients and to develop surveillance indicators common to the three participating MCOs. Using 1993 data, the availability, specifications, and limitations of various surveillance indicators weredetermined. RESEARCH DESIGN AND METHODS Computerized inpatient, pharmacy, outpatient, and laboratory records were used to develop an algorithm to identify diabetes patients and to develop surveillance indicators common to the three participating MCOs. Using 1993 data, the availability, specifications, and limitations of various surveillance indicators were determined RESULTS An extensive set of diabetes surveillance indicators was identified from the four sources of data. Consistent data specifications across MCOs needed to consider variation in the type of data collected, a lack of documentation on level of coverage, differences in codingdata, and different models of health care delivery. A total of 16,363 diabetes patients were identified. The age-adjusted prevalence of diabetes ranged from 24 to 29 per 1,000 enrollees. Approximately one-third of patients with diabetes (32–34%) were taking insulin. The majority had one or more visits to a primary care physician during the year (72–;94%). Visits to specialists were less frequent. Ophthalmologists and optometrists were the most commonly used specialists: 29–;60% of the patients with diabetes at the three MCOs had visited an ophthalmologist or optometrist. About one-fifth had an overnight hospital stay during the year. CONCLUSIONS This diabetes surveillance system is a useful tool for MCOs to track trends in prevalence of diabetes, use of health services, and delivery of preventive care to individuals with diabetes. This system may also be useful for health care planning and for assessing use changes after new developments in diabetes care or new quality management initiatives.

Primary care provider practices and beliefs related to cervical cancer screening with the HPV test in Federally Qualified Health Centers.

When background and other characteristics are controlled for, older adolescents who rear their children are as likely as those who place them for adoption to complete high school. However, relinquishers are more likely to complete vocational training and have higher educational aspirations. Further, relinquishers are more likely to delay marriage, to be employed six and 12 months after the birth and to live in higher income households than are child rearers. Child rearers are more likely to become pregnant again sooner and to resolve subsequent pregnancies by abortion. Adolescents who relinquish their children do not suffer more negative psychological consequences than do those who raise their children. Overall, both groups indicated very high levels of satisfaction with their decision to relinquish or to rear, although relinquishers were slightly less satisfied with their decision than were child rearers. The study sample consisted of 123 child rearers and 146 relinquishers who had attended a pregnancy-counseling program affiliated with a large adoption agency that practices open adoption. Hence, the findings are limited to a select sample and should not be generalized beyond adolescents who participate in a similar program.

Colorectal Cancer Screening: Estimated Future Colonoscopy Need and Current Volume and Capacity

Although surgical proficiency is essential to the immediate outcome of transplantation, long-term success depends upon how adequately the transplantation recipient is managed. Immunosuppression, the most critical aspect of after care, is subject to wide variation. In January 1990, a survey was sent to the directors of all transplant programs in the United States performing one or more kidney, heart, liver, heart-lung, or pancreas transplant in 1988. Detailed data were obtained on both the drugs and methods used for induction and maintenance immunosuppression, as well as the treatment of rejection. Each program director was asked to rank each immunosuppressive approach according to its perceived impact on patient outcomes. Over 85% of all eligible program directors completed the survey. There is no evidence of survey respondent bias. The use of polyclonal and monoclonal agents for induction immunosuppression was favored most by pancreas program directors (72–76%). These agents were least preferred by liver transplant programs (35–37%). About half of kidney, heart, and heart-lung program directors preferred these agents. Triple-drug therapy consisting of CsA, PRED, and AZA was considered the most preferable maintenance protocol for all transplants (i.e., kidney, 89%; heart, 94%; liver, 88%; heart-lung, 86%; pancreas, 96%). Either i.v. steroids or OKT3 were regarded as the preferred approaches for the treatment of acute or resistant rejection. Finally, the acceptability of outpatient treatment of rejection varied by transplant type (i.e., kidney, 9%; heart, 58%; liver, 5%; heart-lung, 29%; pancreas, 8%). Although there are similarities in the ratings of various aspects of immunosuppressive therapy, there are important differences. This information

Cost Effectiveness of ACE Inhibitor Treatment for Patients with Type 1 Diabetes Mellitus

OBJECTIVE Cervical cancer screening using the human papillomavirus (HPV) test and Pap test together (co-testing) is an option for average-risk women ≥ 30 years of age. With normal co-test results, screening intervals can be extended. The study objective is to assess primary care provider practices, beliefs, facilitators and barriers to using the co-test and extending screening intervals among low-income women. METHOD Data were collected from 98 providers in 15 Federally Qualified Health Center (FQHC) clinics in Illinois between August 2009 and March 2010 using a cross-sectional survey. RESULTS 39% of providers reported using the co-test, and 25% would recommend a three-year screening interval for women with normal co-test results. Providers perceived greater encouragement for co-testing than for extending screening intervals with a normal co-test result. Barriers to extending screening intervals included concerns about patients not returning annually for other screening tests (77%), patient concerns about missing cancer (62%), and liability (52%). CONCLUSION Among FQHC providers in Illinois, few administered the co-test for screening and recommended appropriate intervals, possibly due to concerns over loss to follow-up and liability. Education regarding harms of too-frequent screening and false positives may be necessary to balance barriers to extending screening intervals.

Overview of the CDC Cervical Cancer (Cx3) Study: An Educational Intervention of HPV Testing for Cervical Cancer Screening

In 2014, a national campaign was launched to increase colorectal cancer (CRC) screening rates in the United States to 80% by 2018; it is unknown whether there is sufficient colonoscopy capacity to reach this goal. This study estimated the number of colonoscopies needed to screen 80% of the eligible population with fecal immunochemical testing (FIT) or colonoscopy and determined whether there was sufficient colonoscopy capacity to meet the need.

The longitudinal cost of end-stage renal disease in persons with diabetes in the United States

AbstractObjective: Current guidelines recommend treating patients with type 1 diabetes mellitus with ACE inhibitors after the onset of microalbuminuria. Recent clinical trials have shown ACE inhibitors can affect the development of nephropathy when initiated prior to the onset of microalbuminuria. Our objective is to examine the cost effectiveness of treating adults aged over 20 years with an ACE inhibitor (captopril) immediately following diagnosis of type 1 diabetes versus treating them after the onset of microalbuminuria. Design: Using a semi-Markov model, we calculated four main outcome measures: lifetime direct medical costs (discounted), QALYs, cumulative incidence of end-stage renal disease (ESRD), and number of days of ESRD over a lifetime. Medical costs are in 1999 US dollars. Setting: All analyses were from the viewpoint of a single US payer responsible for all direct medical costs, including screening for microalbuminuria, ACE inhibitor treatment (captopril), management of major diabetic complications, and routine annual medical costs not specific to diabetes. Methods: We applied the model to a hypothetical cohort of 10 000 persons newly diagnosed with type 1 diabetes. Distribution of sex and race/ethnicity within the cohort is representative of the general US population. Results: We estimated that the incremental cost of early use of captopril for the average adult with type 1 diabetes is