Outcomes after outpatient endoscopy: Can administrative data tell the whole story?

Historically, endoscopy has been performed almost exclusively in hospital settings. However, as experience and indications have grown, endoscopy has been increasingly performed in outpatient facilities separate from hospitals. This trend was earlier and more prevalent internationally in jurisdictions with private or partially private healthcare systems but has gained a foothold even in single-payer/universal healthcare systems such as in Canada.1Murthy S.K. Benchimol E.I. Tinmouth J. et al.Temporal trends in postcolonoscopy colorectal cancer rates in 50- to 74-year-old persons: a population-based study.Gastrointest Endosc. 2018; 87: 1324-1334Abstract Full Text Full Text PDF PubMed Scopus (15) Google Scholar These facilities can range from being extensions of a hospital system, to a collaboration with a hospital, to completely separate entities. Depending on the jurisdiction, these facilities may be regulated and/or accredited by a healthcare authority and typically have regulatory or self-imposed restriction criteria on their patient population and procedural complexity in efforts to minimize adverse events and need for hospital transport. These facilities offer additional regional endoscopy access for patients, and they may be a response to payer pressure for increased productivity at lower cost. A reduction in the aggregate cost per procedure is typically achieved through increasing the number of procedures performed within the same cost envelope and/or decreasing the complexity/equipment costs of the procedures. In addition, these facilities may be a response to more intangible factors such as provider freedom, autonomy, or move to higher-volume lower-risk procedures, to name a few. In this issue of Gastrointestinal Endoscopy, Lin et al2Lin M.-Y. Mishra G. Ellison J. et al.Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort study.Gastrointestinal Endosc. 2022; 95: 1088-1097.e17Abstract Full Text Full Text PDF Scopus (1) Google Scholar present the results of a statewide cohort study of nearly 2 million procedures comparing the adverse event rates in 21 ambulatory surgical centers (ASCs) versus 62 hospital outpatient departments (HOPDs). The authors hypothesized that gross adverse event rates should be lower in ASC because of the selection of lower-complexity patients and procedures but that adjusted adverse event rates would be similar. The authors used the Massachusetts All-Payer Claims Database and Medicare fee-for-service claims to identify patients of all ages undergoing a colonoscopy or EGD in either an ASC or an HOPD in Massachusetts from 2014 to 2017. They used propensity-score matching to build a matched retrospective cohort of a total of 721,788 of the nearly 2 million procedures. Logistic regression was used to build the statistical model for risk ratios. The primary outcome was all-cause unplanned hospital visits occurring within 7 and 30 days after the index GI endoscopic procedure. This was used as a surrogate of a range of postprocedure events, including “adverse events of surgical procedures,” GI bleeding, perforation, abdominal pain, and cardiopulmonary adverse events. These specific outcomes were not presented in the main report but were provided in supplemental appendices. The authors used the Centers for Medicare & Medicaid Services readmission algorithm to minimize the inclusion of planned admissions. Not surprisingly, the authors found that patients attending HOPDs were sicker and differed from those using ASC on most demographic and clinical characteristics. HOPD patients also had a higher proportion of Medicaid coverage. The authors found that lower unadjusted unplanned hospital visit rates in ASCs compared with HOPD procedures, and these rates remained lower for ASCs after adjustment for patient and procedural factors. For the entire cohort, the 7-day unplanned hospital visit rates were 10.7, 18.3, and 38.9 per 1000 procedures for screening colonoscopy, nonscreening colonoscopy, and EGD, respectively. For procedures performed in ASC, the unadjusted rates were 8.5, 13.2, and 9.3 per 1000 procedures, respectively (Table 2 in Lin et al2Lin M.-Y. Mishra G. Ellison J. et al.Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort study.Gastrointestinal Endosc. 2022; 95: 1088-1097.e17Abstract Full Text Full Text PDF Scopus (1) Google Scholar). There was considerable variability in these estimates by facility. For example, the estimates ranged from 4.9 to 49.5 per 1000 nonscreening colonoscopies. There was greater variability in 30-day hospital visit rates, which were reported to be 2 to 3 times greater than the 7-day measures. In the matched adjusted analyses, ASC procedures compared with HOPD procedures were associated with a lower odds ratio of 7-day and 30-day unplanned hospital visits (adjusted odds ratio [aOR], 0.88; 95% confidence interval [CI], 0.79-0.98 for screening colonoscopy; aOR, 0.84; 95% CI, 0.75-0.94 for nonscreening colonoscopy; aOR, 0.57; 95% CI, 0.50-0.65 for EGD for visits within 7 days). Poor health status, increasing comorbidities, and procedural complexity were associated with higher odds of unplanned hospital visits. Administrative data can be extremely useful for studying outcomes in large sets of patients when prospective data are unavailable or challenging to obtain. However, the coding procedures require strict standardization, which inherently will remove important characteristics that can be recovered only with careful chart review. A common example of this is polypectomy codes. Typically, a claims database would have a code for a standard polyp (eg, >3 mm) and another code for a large polyp (>3 cm).3Ontario Schedule of benefits. Available at: https://www.health.gov.on.ca/en/pro/programs/ohip/sob/physserv/sob_master.pdf. Pages S7, S11, S19. Accessed March 28, 2022.Google Scholar Therefore, 2 procedures with a standard polyp code may be vastly different and would not be distinguishable except with manual chart review. Clinical patient characteristics in administrative data are dependent on what is recorded in the medical record and how the record is subsequently interpreted and coded for the database. A major limitation of claims-based studies is understanding what is a procedure-related adverse event versus what is not. Unplanned hospital visits could result from 1 of 3 mechanisms: (1) as a result of the procedure (an adverse event), (2) as a result of the underlying condition being investigated (for example, ongoing pain not caused by the procedure), and (3) a visit unrelated to 1 or 2 (a comorbidity or another illness). In the case of screening colonoscopy, where there is no underlying condition, the unplanned hospital visit rate is likely a more accurate representation of procedural adverse events. Whereas it is somewhat higher, the reported by Lin et al2Lin M.-Y. Mishra G. Ellison J. et al.Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort study.Gastrointestinal Endosc. 2022; 95: 1088-1097.e17Abstract Full Text Full Text PDF Scopus (1) Google Scholar is similar to the rate of 6.45 per 1000 at a large Canadian colon cancer screening center reported rate by our group4Hilsden R.J. Dube C. Heitman S.J. et al.The association of colonoscopy quality indicators with the detection of screen-relevant lesions, adverse events, and postcolonoscopy cancers in an asymptomatic Canadian colorectal cancer screening population.Gastrointest Endosc. 2015; 82: 887-894Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar and a rate of 3.18 per 1000 for directly or potentially related to colonoscopy reported by Ko et al.5Ko C. Riffle S. Michaels L. et al.Serious complications within 30 days of screening and surveillance colonoscopy are uncommon.Clin Gastroenterol Hepatol. 2010; 8: 166-173Abstract Full Text Full Text PDF PubMed Scopus (174) Google Scholar We and others have found that most hospital visits related to the procedure occurred within the first 7 days and that many more events within the first week were procedure-related compared to events after 14 days.6Hilsden R.J. Maxwell C.M. Forbes N. et al.Development of a definition and rules for causal attribution of post-colonoscopy bleeding.PLoS One. 2020; 15e0235902Crossref PubMed Scopus (3) Google Scholar, 7Grossberg L.B. Vodonos A. Papamichael K. et al.Predictors of post-colonoscopy emergency department use.Gastrointest Endosc. 2018; 87: 517-525Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar, 8Dubé C. Rabeneck L. Seven-day postcolonoscopy emergency department visits: what do they really measure?.Gastrointest Endosc. 2018; 87: 526-528Abstract Full Text Full Text PDF PubMed Scopus (1) Google Scholar This suggests that the greater variability observed in the 30-day rates observed by Lin et al2Lin M.-Y. Mishra G. Ellison J. et al.Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort study.Gastrointestinal Endosc. 2022; 95: 1088-1097.e17Abstract Full Text Full Text PDF Scopus (1) Google Scholar is at least in part due to contamination with visits resulting from patients’ underlying conditions or unrelated illnesses. Similarly, the rates for nonscreening colonoscopy and EGD are likely overrepresented with visits due to the underlying condition, especially in the 30-day estimates. This is especially true for the unexpected finding of higher adverse event rates for EGD over colonoscopy. For EGD, “adverse events of surgical procedures or medical care” did not even feature in the top 5 reasons for hospital visits (Supplementary Table 5 in Lin et al2Lin M.-Y. Mishra G. Ellison J. et al.Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort study.Gastrointestinal Endosc. 2022; 95: 1088-1097.e17Abstract Full Text Full Text PDF Scopus (1) Google Scholar). Likewise, for nonscreening colonoscopy, chest pain, abdominal pain, and diverticulosis/diverticulitis accounted for 18.5% of hospital visits, suggesting that hospital visits may have been contaminated by patients presenting for preexisting symptoms. Last, the HOPD patient cohort appeared to have more comorbidities despite matching, which could have resulted in higher rates of hospital visits due to their comorbidities, unrelated illnesses, or other factors such as prior healthcare-seeking behavior (Supplementary Table 12 in Lin et al2Lin M.-Y. Mishra G. Ellison J. et al.Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort study.Gastrointestinal Endosc. 2022; 95: 1088-1097.e17Abstract Full Text Full Text PDF Scopus (1) Google Scholar). The use of 7-day hospital visit rates has also been questioned because they appear to correlate better with patients’ prior hospital visit rates than to the endoscopic procedure.7Grossberg L.B. Vodonos A. Papamichael K. et al.Predictors of post-colonoscopy emergency department use.Gastrointest Endosc. 2018; 87: 517-525Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar Aggregate unplanned hospital visits are not an ideal surrogate for procedure-related adverse events and may obscure true signals in the data. For example, in the current study, for screening and nonscreening colonoscopy, the rates for unplanned hospital visits for “adverse events of surgical procedures or medical care” were actually numerically higher in ASCs versus HOPDs in the matched cohort (13.4% vs 8.7%, screening colonoscopy; 11.1% vs 7.6%, nonscreening colonoscopy) (Supplementary Table 5 in Lin et al2Lin M.-Y. Mishra G. Ellison J. et al.Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort study.Gastrointestinal Endosc. 2022; 95: 1088-1097.e17Abstract Full Text Full Text PDF Scopus (1) Google Scholar). Such a signal from the analysis should trigger a more comprehensive patient-level, data review in a follow-up study. Ultimately, the authors present evidence that suggests endoscopy can be performed safely in ASCs with postprocedural unplanned hospital visit rates that are similar to the rates for like procedures performed in HOPDs. However, we would caution against interpreting the current study results to suggest that hospital visit rates are lower in ASCs than in HOPDs for identical procedures and patients as reported in this study. In the end, to truly understand differences in procedure-related adverse event rates, one needs more granular data on the patients, the procedures, and the events than are offered with administrative data. This could be achieved with studies using dedicated prospective data collection or comprehensive chart reviews of retrospective data. Endoscopy quality and safety go hand in hand and should not be different when the same procedure is performed in one setting versus another. Endoscopy units, regardless of setting, can further assure high-quality care and safety through the adoption of validated quality monitoring and improvement paradigms such as the Global Rating Scale or the American Society for Gastrointestinal Endoscopy quality guidelines.9Hilsden R.J. Rostom A. Dubé C. et al.Development and implementation of a comprehensive quality assurance program at a community endoscopy facility.Can J Gastroenterol. 2011; 25: 547-554Crossref PubMed Scopus (12) Google Scholar, 10MacIntosh D. Dubé C. Hollingworth R. et al.The endoscopy Global Rating Scale–Canada: development and implementation of a quality improvement tool.Can J Gastroenterol. 2013; 27: 74-82Crossref PubMed Scopus (31) Google Scholar, 11Day L. Cohen J. Greenwald D. et al.ASGE Endoscopy Unit Quality Indicator TaskforceQuality indicators for gastrointestinal endoscopy units.VideoGIE. 2017; 2: 119-140Abstract Full Text Full Text PDF PubMed Scopus (20) Google Scholar, 12Chapman F.J. Cohen J. et al.ASGE Ensuring Safety in the Gastrointestinal Endoscopy Unit Task Force; Calderwood AHGuidelines for safety in the gastrointestinal endoscopy unit.Gastrointest Endosc. 2014; 79: 363-372Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar All authors disclosed no financial relationships. Differences in patient outcomes after outpatient GI endoscopy across settings: a statewide matched cohort studyGastrointestinal EndoscopyVol. 95Issue 6PreviewOutpatient GI endoscopy has been shifting from hospital outpatient departments (HOPDs) to ambulatory surgery centers (ASCs) in recent years. However, evidence on whether patient outcomes after endoscopic procedures are comparable across settings is limited. This study compares the incidence of unplanned hospital visits after GI endoscopy performed in ASCs versus HOPDs. Full-Text PDF Open Access

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.