Working with a group of key stakeholders, the authors developed an episode-based resource use measure focused on the use of colonoscopy. This measure is intended to identify differences in health care resource use in a short time frame surrounding the colonoscopy. The ultimate intent in the development of this measure was to pair it with a measure of quality so that both the cost and quality of care can be evaluated together. In initial testing, the authors found the use of general anesthesia with colonoscopy to be associated with higher episode costs. Eventually, when paired with quality measures, it is hoped this measure will provide actionable information for health care payers and providers to more efficiently provide colonoscopy services without compromising quality.
Numerous studies have indicated that the United States spends significantly more per person on health care than any other nation in the world. Additionally, research has documented significant variation in spending by provider and by region in the United States and that this variation often has little or no correlation with the quality of care provided or with patient outcomes. Although reducing health care spending is a central goal of the health care reform debate, clear evidence on the best ways to do so remains out of reach.
Although documentation of variability in the overall costs of care at regional levels points out that inefficiencies exist in the health care system, it does not provide actionable information on what may be the underlying cause of the differences and how these differences can be reduced. One potential solution is to focus on episode-based resource use and costs so that differences within a particular clinical area can be examined and areas in need of action can be identified. Moreover, episode-based measures may be combined with quality of care measures to provide some insights in identifying efficient care in which quality is high and costs are low. Such information would allow all parties involved (consumers, purchasers, and providers) to better understand how treatment decisions affect the costs and quality of their care. Data gathered from such analyses have the potential to provide clear and actionable information on what components of care can (or should) be reduced and what components of care can (or should) be increased, thereby helping to reduce spending while at the same time maintaining or even improving clinical quality and outcomes.
Ideally, in order for a given condition or procedure to be a candidate for an episode-based measure of health care costs, a clear body of evidence that supports the relevant clinical management and treatment decisions should be readily available. The use of colonoscopy in colon cancer screening is one of these areas highly suitable for such episode-based measure development. We have developed an episode-based measure of costs of care associated with screening colonoscopy, and our subsequent analytic findings of the measure suggest that it can serve as a helpful tool for identifying, and potentially addressing, unwarranted variability in resource use related to the performance of such procedures.
Colorectal cancer is one of the most commonly diagnosed cancers in the United States. In 2005, more than 140,000 men and women were diagnosed with colorectal cancer, and approximately 52,000 died from of the condition. Both the Institute of Medicine and the Ambulatory Care Quality Alliance (AQA) have identified colon cancer as 1 of 20 condition-specific priority areas in need of quality improvement, based on its relevance to a considerable volume of patients, its impact on those patients, and the perception of opportunity to significantly improve the quality and efficiency of related care. The complete list of priority condition-specific areas is included as follows.
- •
Acute Myocardial Infarction
- •
Angina/Coronary Artery Disease
- •
Asthma
- •
Breast Cancer
- •
Bronchitis
- •
Chronic Obstructive Pulmonary Disease
- •
Colon Cancer
- •
Congestive Heart Failure
- •
Depression
- •
Diabetes
- •
Hiatal Hernia (Gastroesophageal Reflux Disease)
- •
Hip Fracture
- •
Hypertension
- •
Hysterectomy
- •
Low Back Pain
- •
Osteoarthritis
- •
Pneumonia
- •
Prostate Cancer
- •
Sinusitis
- •
Stroke.
Colorectal cancer screening has been shown to reduce colorectal cancer mortality by as much as 60%. Although there are a variety of ways to screen for colorectal cancer, the most popular method used today is colonoscopy. This procedure involves the insertion through the rectum of a flexible videoscope, which is then advanced proximally through the entire length of the colon to search for the presence of polyps. Colonoscopy is the preferred colorectal cancer screening strategy of both the American College of Gastroenterology (ACG) and the American Society of Colon and Rectal Surgeons (ASCRS), receiving a Grade 1B recommendation in the ACG’s most recent guidelines (issued in 2008). Colon cancer screening is similarly recommended by the US Preventive Services Task Force and has also been identified as a priority area in other national initiatives, including the Health Resources and Services Administration’s (HRSA) Health Disparities Collaboratives and the Centers for Medicare and Medicaid Services’ (CMS) Quality Improvement Program.
Although the role of colonoscopy in detecting and preventing colon cancer is clear, concerns have been raised in recent years about the overall rising costs of the procedure. These concerns are in part based on the increasing total volume of colonoscopy procedures performed as well as the increasing costs of each individual procedure. In 2003, for example, 30% of eligible women and 32% of eligible men 50 years and older had undergone the procedure. The rising costs of each procedure may largely be attributable to increasing costs of ancillary resources that are used. For example, because patient discomfort during the procedure can be considerable, some sort of sedation or anesthesia is typically administered. However, the type of sedation given, whether or not more complete anesthesia should be used, and whether or not sedation is even necessary at all in every circumstance is of some debate. As a result, considerable individual provider discretion is the norm. Furthermore, the procedure has some inherently associated potential complications (eg, bleeding and bowel perforation), and the potential for these complications to occur may also vary depending on the level of sedation. Whereas procedures performed with sedation have higher risks of respiratory depression, falls, and other sedation-related complications, those performed without sedation have higher failure rates in part because of patient discomfort.
Measuring resource use and costs of care
Alternative methodologies exist that can be used to measure the resource use associated with a colonoscopic examination and individual health care costs in general. The 2 primary approaches are per-capita measurement and per-episode measurement. Per-capita measurement captures the cumulative health care costs for a given population. Examples of this methodology include measures of total costs per member per month (PMPM) and measures of service use per 1000 patients per year. Although this methodology is relatively easy to implement and interpret, the measures themselves are population-level measures, and accurately establishing and assigning accountability for such population-level measures can be difficult. One reason for this difficulty is the dispersed nature of the medical care that many patients receive. Researchers have found that because many patients see multiple providers for multiple conditions over the course of a given year, assigning meaningful accountability for the total costs of their care can be problematic.
Per-episode measurement quantifies the services involved in the diagnosis, management, and treatment of unique clinical conditions. These measures can capture the total costs associated with any acute or chronic condition of interest. Episode-based measures can also be more focused and precise—as in the case of the colonoscopy measure we describe later in this article. This particular episode-based measure focuses on the patient’s preparation before the colonoscopy, the procedure itself, and any related complications following the procedure. Any postprocedure patient management or treatment that may be related to a newly established diagnosis, however, would not be included here; such patient management or treatment could potentially be included in other measurement efforts. Although significantly more complex to develop than per-capita measures, episode-based measures have the advantage of increased clinical specificity and are potentially much more actionable for providers and consumers.
Although commercial vendors currently offer tools that rely on per-episode measurement structures to generate estimates of physician performance based on cost (most prominently Ingenix’s Episode Treatment Groups and Thomson-Reuters’ Medical Episode Grouper ), early efforts to implement these tools have experienced only limited success. Key issues affecting these implementation efforts have included a lack of transparency in the measurement methodologies, inconsistent or ineffective communication with patients and providers during the implementation process, and provider resistance to cost-of-care measurement in any form.
The High Value Health Care (HVHC) Project, funded by the Robert Wood Johnson Foundation and overseen by the Quality Alliance Steering Committee (QASC), is working to make valid, timely, and consistent information about the quality and cost of health care widely available in the United States through furthering the development and use of a performance measurement infrastructure. One component of this effort specifically focuses on the development of a fully transparent set of cost-of-care measures developed with the input and clinical guidance of key stakeholders (including practicing physicians). The goal of structuring the measure development process in this way is to alleviate many of the concerns providers have expressed about other, less transparent, cost-of-care measurement algorithms, such as the proprietary efforts cited previously. a
a More information regarding the HVHC cost-of-care measure development effort can be found here: http://www.healthqualityalliance.org/hvhc-project/cost-care-measurement-development .
Under this component of the HVHC project called Characterizing Episodes and Cost of Care (C3), the American Board of Medical Specialties Research and Education Foundation (ABMS REF), in conjunction with The Brookings Institution, undertook the development of episode-based cost-of-care measures for 12 of the most prevalent and important acute and chronic conditions in the United States that were identified by the AQA.
Here, we discuss the process by which an episode-based measure of the costs of care associated with colonoscopy has been developed as part of the C3 project and show some of the effort’s preliminary analytic findings.
Building an episode of care for colonoscopy
For each of the 12 high-priority conditions included in the HVHC Project, a group of expert clinicians and other stakeholders was convened for a 2-day in-person meeting followed by a series of teleconference meetings. During the in-person meeting, the workgroup focused initially on conceptually defining one or more important measures for the condition, reaching a broad consensus on the cohort definition (including which patients should be excluded from the measure). Then the work group sought to identify the services and resource use that were “related” to the clinical condition of interest for each measure. In this context, “related” services and resource use were defined as all of the medical care provided with the intent of treating or managing the clinical condition of interest as well as all of the care involved in the management of any resulting complications. Notably, this definition does not distinguish any care that is appropriate from any care that is not, nor does it distinguish any care that is in compliance with generally accepted clinical treatment guidelines from care that is not. Relative to the colonoscopy episode, “related” resource use consists of all the care pertinent to the performance of the examination (eg, the procedure itself, sedatives and other medications, attendant supervision), any immediate preparations for the procedure (eg, the bowel preparations, other medications), and any pertinent conditions that may arise in the immediate postprocedure period (eg, bowel perforations, bleeding, repeat colonoscopies).
Following the in-person meeting, the concepts were translated into detailed measure specifications for further review by the clinician work groups. The measure specifications were then developed into a series of computer algorithms and tested using a large administrative claims dataset, benchmark statistics from national research organizations, and pertinent information from the clinical literature. b
b The dataset used for these analyses was the MarketScan Commercial Claims & Encounters Database provided by Thomson Reuters (Healthcare), Inc. The MarketScan data contain claims information for a large population of individuals aged 0 to 64 who were enrolled in a commercial insurance plan during the calendar year 2006 or 2007. In total, the data reflect the health care experience of approximately 15 million covered lives per year. Although all regions of the United States are represented, patients in the database are disproportionately from the South.
The variability of costs across regions and provider specialties associated with the episode or episodes developed through this process was examined to determine whether each measure was effective in identifying unwarranted variation in costs (ie, variation not attributable to underlying variation in patient complexity or morbidity).Through this process, a measure was developed that focuses on variation in resource use observed in the 22-day period surrounding a screening colonoscopic examination with the clinical input of the colon cancer clinical work group; primarily methodological input was also provided by the C3 Project’s Technical Advisory Committee and the QASC’s Episodes Workgroup. The measure includes the resources used during a 7-day period preceding the colonoscopy, those used on the day of the examination, and those used during a 14-day period following the procedure. Members of the work group anticipated that some variability might be observed in the measure’s resource use across episodes both as a result of the type of colonoscopy performed (eg, no biopsies, biopsies or polypectomies) and as a result of the types of ancillary services used (eg, no sedation, conscious sedation, or general anesthesia). Additional variation might be seen if there were complications (eg, antibiotics and other medications, lab tests, radiographic examinations, corrective surgical procedures) or if there was a need to repeat the colonoscopy examination itself. Work group members felt that such complication-related resource use would most likely be captured during the 2-week period immediately following the procedure.
Building an episode of care for colonoscopy
For each of the 12 high-priority conditions included in the HVHC Project, a group of expert clinicians and other stakeholders was convened for a 2-day in-person meeting followed by a series of teleconference meetings. During the in-person meeting, the workgroup focused initially on conceptually defining one or more important measures for the condition, reaching a broad consensus on the cohort definition (including which patients should be excluded from the measure). Then the work group sought to identify the services and resource use that were “related” to the clinical condition of interest for each measure. In this context, “related” services and resource use were defined as all of the medical care provided with the intent of treating or managing the clinical condition of interest as well as all of the care involved in the management of any resulting complications. Notably, this definition does not distinguish any care that is appropriate from any care that is not, nor does it distinguish any care that is in compliance with generally accepted clinical treatment guidelines from care that is not. Relative to the colonoscopy episode, “related” resource use consists of all the care pertinent to the performance of the examination (eg, the procedure itself, sedatives and other medications, attendant supervision), any immediate preparations for the procedure (eg, the bowel preparations, other medications), and any pertinent conditions that may arise in the immediate postprocedure period (eg, bowel perforations, bleeding, repeat colonoscopies).
Following the in-person meeting, the concepts were translated into detailed measure specifications for further review by the clinician work groups. The measure specifications were then developed into a series of computer algorithms and tested using a large administrative claims dataset, benchmark statistics from national research organizations, and pertinent information from the clinical literature. b
b The dataset used for these analyses was the MarketScan Commercial Claims & Encounters Database provided by Thomson Reuters (Healthcare), Inc. The MarketScan data contain claims information for a large population of individuals aged 0 to 64 who were enrolled in a commercial insurance plan during the calendar year 2006 or 2007. In total, the data reflect the health care experience of approximately 15 million covered lives per year. Although all regions of the United States are represented, patients in the database are disproportionately from the South.
The variability of costs across regions and provider specialties associated with the episode or episodes developed through this process was examined to determine whether each measure was effective in identifying unwarranted variation in costs (ie, variation not attributable to underlying variation in patient complexity or morbidity).Through this process, a measure was developed that focuses on variation in resource use observed in the 22-day period surrounding a screening colonoscopic examination with the clinical input of the colon cancer clinical work group; primarily methodological input was also provided by the C3 Project’s Technical Advisory Committee and the QASC’s Episodes Workgroup. The measure includes the resources used during a 7-day period preceding the colonoscopy, those used on the day of the examination, and those used during a 14-day period following the procedure. Members of the work group anticipated that some variability might be observed in the measure’s resource use across episodes both as a result of the type of colonoscopy performed (eg, no biopsies, biopsies or polypectomies) and as a result of the types of ancillary services used (eg, no sedation, conscious sedation, or general anesthesia). Additional variation might be seen if there were complications (eg, antibiotics and other medications, lab tests, radiographic examinations, corrective surgical procedures) or if there was a need to repeat the colonoscopy examination itself. Work group members felt that such complication-related resource use would most likely be captured during the 2-week period immediately following the procedure.
Colonoscopy episode-of-care cohort definition
Accurately defining the population of interest is critical to identifying meaningful variation in resource use for any condition’s treatment. Although it may seem a straightforward choice to capture all colonoscopies in a given year, this approach can lead to the inclusion of individuals who receive colonoscopies for reasons other than the early detection of colon cancer. Doing so could introduce confounding variability in resource use into the measure. For this and other related reasons, the clinicians participating in the measure development elected to define the colonoscopy measure’s eligible population as follows.
Inclusion Criteria
Patients are included in the measure if they had a colonoscopy billed using any of the codes listed in Table 1 during the 22-day period covered by the episode and if they are 40 years or older at the time of the procedure. c
c We acknowledge that the existing clinical guidelines for colonoscopy recommend screening only for those aged 50 years and older (45 years for African Americans); however, it was the opinion of the clinical expert panel that there are many patients who undergo screening before age 50, usually because of a family history of colon cancer. Additionally, resource use for colonoscopy was not expected to differ significantly by age.
As Table 1 illustrates, CPT code 45378 (Colonoscopy, flexible, proximal to splenic flexure; diagnostic, with or without collection of specimen[s] by brushing or washing, with or without colon decompression [separate procedure]) accounts for 46% of all colonoscopies in our sample, followed by CPT code 45380 (Colonoscopy, with biopsy, single or multiple) which represents 28% of colonoscopies in our sample. Of the remaining types of colonoscopy, only CPT 45385 (Colonoscopy with removal of tumor[s], polyp[s], or other lesion[s] by snare techniques) accounts for more than 10% of cases.Related posts:
Reducing Infection Risk in Colonoscopy
Sedation Issues in Quality Colonoscopy
Complications of Colonoscopy: Magnitude and Management
Colonoscopy Reports and Current State of Performance Measures
Advanced Systems to Assess Colonoscopy
The Use of Databases and Registries to Enhance Colonoscopy Quality
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
