Area differences in healthcare utilization have been documented since an area variation analysis of tonsillectomy rates was published back in 1938. Besides the research literature, geographic variation in U.S. healthcare also has been documented extensively for the Medicare population in the Dartmouth Atlas. The Institute of Medicine also released a series of reports on geographic variation in healthcare spending in early 2013. General consensus from these studies is that significant regional difference in various health services is observed after adjusting for available patient and payer differences.
Regional variation in healthcare utilization is greatest among procedures or diagnoses with variation in standard practices. Syncope, for example, can be treated efficiently on an ambulatory basis or in the inpatient setting up to 1 day. It largely depends on physicians’ preference as to whether to treat a syncope patient outside the hospital or keep the patient overnight. The chart below illustrates the regional difference in Emergency Room and/or Observation Care (“ER/OBS”) utilization and inpatient admissions for syncope between four New York and New Jersey plans and four California plans. Data for all plans are limited to commercial members, and utilization rates are adjusted to reflect standard commercial demographics. Compared to the commercial average (regression line), the four California plans stay at the lower ends of both ER/OBS visits per 1000 and inpatient admission via ER/OBS per 100. On the contrary, the four New York/New Jersey plans not only had a higher than average rates of syncope ER/OBS visits but also admitted a higher percentage of syncope patients via ER/OBS to the inpatient setting.
The difference in ER/OBS visits rate between the California plans and the New York/New Jersey plans for syncope perhaps reflects not only where syncope patients seek care in different areas but also the availability of primary care sources in each area. It may also reflect different patient understandings of the appropriate role of Emergency Department care. However, the difference in the syncope ER/OBS admission ratio (syncope admissions via ER/OBS per 1000 divided by syncope ER/OBS visits per 1000) reflects more on differences physician decision-making. For this condition, physicians in New York/New Jersey admitted ER/OBS syncope patients to the inpatient setting much more often than those in California did (9.6% – 16.6% vs. 6.5% – 8.4%). Even assuming that the same diagnostic work-up takes place in either the inpatient or the outpatient setting, the added cost of accommodating this much higher percentage of patients overnight in the hospital can be substantial.
The methodology embedded in the MedInsight Guideline Analytics software solution was used to create the analysis for this article, e.g. the definition of syncope in medical claims data and the means to count admits, observation, and ER cases.
*This article is a guest post, written collaboratively by John Ninomiya, Ph.D, FSA , Director, Epidemiology & Health Services Research, MCG Health and Frances Lee, Ph. D., Senior Clinical Benchmarking Advisor, MCG Health.
Observation care continues to be a hot topic in the news for both Medicare and Commercial health plans in the United States today. One aspect of observation care is the lack of a clear definition of what constitutes an observation case. In the news the primary focus on observation case definition revolves around when an observation case should be considered an inpatient stay. For example, on July 30th, an investigation conducted by the Department of Health and Human Services Inspector General found both Medicare officials and hospitals are struggling to fully understand the difference between observation and inpatient status. In fact this report noted that the six of the top 10 reasons for observation care were also among the 10 most frequent reasons for a short inpatient hospital stay of one night or less, http://www.kaiserhealthnews.org/Stories/2013/July/30/IG-report-observation-care.aspx. In June of 2013 Premier requested the Centers for Medicare and Medicaid Services define observation care as inpatient after 72 hours of care, http://www.healthdatamanagement.com/news/hospital-long-term-inpatient-prospective-payment-system-46305-1.html.
Additionally, we find there to be a lack of uniformity in the definition for considering observation cases being distinct from an emergency room (ER) case or not. We feel this lack of clarity is important particularly when we have to consider the rising costs and rates of ER usage and the ancillary services being provided in that setting of care.
When Milliman researched how commercial health plans define observation care we found various models. For example we found some contracts to state that when ER services precede an observation stay, the ER services are considered to be incidental to the observation stay and are not separately reimbursed. For other contracts, ER and observation services can both be reimbursed separately. Currently, the Milliman Health Cost Guidelines (HCGs), embodied by an algorithm within the popular Milliman HCG Grouper, have ER services override observations services if they are billed together on the same claim, all costs and services are bundled into the ER case.
In a study of the Milliman normative databases, commercial health plan data in the United States using incurred medial claim data from 2010 and 2011, we found total observation care cases add up to be 40% of all emergency room cases, see the table below.
Furthermore the average allowed unit cost for ER cases with observation services included in the same claim was $3,891.34.
These data points are just the beginning to our exploration to re-evaluate if and when a claim or set of claims should be labeled as observation care or ER. For example should care after 8 hours in the ER shift to becoming an observation case? Are the service units encoded on claims data credible enough to use as hours of observation care in all situations? Again we must note that included in these costs are ancillary services which need to be attributed to either an ER or an observation case. See an earlier blog post on this related topic as well, http://medinsight.wpengine.com/bid/305155/What-is-driving-Emergency-Room-costs, since the increases in services over time have been considered to be a leading driver of cost trends in the ER.
We are now being driven by these questions to work in the coming summer months of 2013, with our clients input, to hopefully add transparency in the process of defining observation care. We look forward to hearing any and all feedback so please email us or add comments to this blog post for the community to discuss.
The CMS diagnostic related groups (DRGs) have undergone numerous refinements since first introduced in the early 1980s, but they remain essentially a tool to support the CMS prospective payment system. What would a grouper that focused on clinical categories rather than payment look like? For the answer, take a tour of our new product – Guideline Analytics.
Guideline Analytics uses standard claims data elements — diagnosis/procedure codes and patient demographics, for example — to assign each inpatient admission to an Optimal Recovery Guideline (ORG) or General Recovery Guideline (GRG) category drawn directly from our acute care content —Inpatient & Surgical Care and General Recovery Care. Each admission is categorized by a principal ORG or GRG code, subsidiary ORG codes, and a severity category based on the comorbidity methodology developed by CMS for its DRG system, complication/comorbidity (CC) and major CC.
The results are easy-to-manipulate data warehouse analytics and/or an Excel spreadsheet report that allows you to analyze performance against optimal outcomes, compare yourself to peer organizations, and identify clinical areas where you can provide care more effectively.
Why not simply use the MS-DRGs to accomplish this analysis? The short answer is that the MS-DRGs were built for a different purpose. For the long answer, let’s say you’re the medical director at a medium-sized health plan. (The data that follows is actual 2010 data from such an organization.) You’re trying to understand how your network is addressing obesity using surgical procedures. Specifically, you want to know how many surgeries are being performed on an inpatient basis that – under optimal conditions – could be performed on an outpatient basis. The DRGs provide three categories for Operating Room Procedures for Obesity. However, two of those categories are for cases with significant complications and comorbidities– unlikely candidates for outpatient surgery, and in practice, only about one-sixth of cases. That leaves one large category (DRG 621) for analysis. Not much granularity.
Guideline Analytics breaks DRG 621 into seven MCG™ guideline categories (see chart below), including S-515 – Gastric Restrictive Procedure without Gastric Bypass by Laparoscopy. Why is S-515 so interesting? According to current medical literature, given optimal conditions, patients can receive this procedure on an outpatient basis. Yet 28% of such procedures are being performed on an inpatient basis at an overall cost of $1.3 million. Is that a reasonable percentage? Should you dig deeper into the data? Guideline Analytics provides risk-adjusted benchmarks – covering different lines of business, different regions of the country, different degrees of medical management, and different delivery systems – so you can assess where that 28% places you compared to peer organizations. Guideline Analytics are linked to MCG™ guideline categories in terms of their foundation only, the Guideline Analytics are not dependent on a client having a license to MCG™ products. Due to the linkage of the Guideline Analytic categories to MCG™ we know that the recent medical evidence on this procedure, S-515, includes five published studies showing the safety of outpatient treatment and four articles describing situations in which inpatient care may be required. This unique combination of statistical data and medical evidence from peer-reviewed sources allows you to decide your next step with confidence.
2010 Inpatient Medical Claims Analysis for Medium-Sized US Health Plan