Identifying Health Care Quality Measures for Medicaid-Eligible Adults: Background Report

4. Results

Workgroup Results

Maternal/Reproductive Health

After discussing the measures against the criteria, the Maternal/Reproductive Health workgroup recommended keeping each of the measures originally posed for the core set, noting that these measures addressed areas of high importance to women and reproductive health, were feasible to report, and aligned well with current programs (including the Initial Core Set of Children's Health Care Quality Measures7). The workgroup noted that, while future measures should tie screenings to outcomes and assess additional issues outside of pregnancy that affect women (e.g., access to care, incontinence due to multiple pregnancies), the measures being recommended for the core set were an important first step of using performance measures for quality improvement. Of the measures newly suggested through public comment, the workgroup recommended bringing one measure forward to a Subcommittee vote: Chlamydia Screening in Women. The workgroup rated this measure high on each criterion and noted its alignment with the Initial Core Set of Children's Health Care Quality Measures (the Initial Core Set of Children's Measures specified only the lower age group of this measure; adding the higher age range means the measure now would be reported in full).

Adult Health

Of the 15 measures initially recommended, this workgroup prioritized 10 measures to be included in the final set, dropping five measures that were duplicative of the types of clinical or chronic conditions other measures being considered addressed and scored lower on the criteria. The workgroup brought forward one measure that was suggested in public comment, Adult Body Mass Index (BMI) Assessment, replacing a similar BMI measure that had been originally recommended for the core set, Preventive Care and Screening: BMI Screening and Follow-Up. The workgroup did not recommend including the remaining 16 newly suggested measures. Several of the measures were narrowly focused on specific clinical processes (e.g., Risk-Adjusted Operative Mortality for Mitral Valve Replacement, Stroke and Stroke Rehabilitation: Anticoagulant Therapy Prescribed for Atrial Fibrillation at Discharge).

Complex Health Care Needs

The Complex Health Care Needs workgroup recommended nine measures for inclusion in the final set. Although the topic areas represented in the measures suggested through public comment were important to Medicaid, many of the measures scored low on multiple criteria and were deemed not ready for wide-scale implementation. Further, although several of the proposed measures assessed the very important topic of care coordination for patients who are hospitalized or transferred across multiple facilities, the workgroup noted that many of these measures were challenged by complex requirements for data collection and excluded target populations (e.g., dually eligible beneficiaries and individuals with long-term care services and supports needs). Many of the measures, for example, required medical record review across time or at more than one site (e.g., Change in Basic Mobility as Measured by the AM-PAC and Medication Reconciliation Post-Discharge). Also, AHRQ's Prevention Quality Indicators, although addressing key clinical quality topics such as dehydration and urinary tract infections, excluded people transferred from nursing home and long-term care facilities. The workgroup concluded that the remaining measures suggested in public comment, though relevant to people with complex health care needs, addressed very narrow clinical conditions, excluded key populations, were difficult to collect at the State level, or were duplicative of other, more highly-rated measures.

Mental Health and Substance Use

After discussing how well the measures fared against the selection criteria, the Mental Health and Substance Use workgroup recommended nine measures for inclusion in the draft core set and decided against bringing forward any of the additional measures suggested in public comment. Workgroup members acknowledged the trade-offs between importance/evidence and feasibility in this topic area. They also noted that interventions or processes that are measurable and more widely available may not adequately capture issues such as safety or outcomes. In general, the workgroup prioritized measures that were broadly applicable to the Medicaid population or to primary care settings. For example, the workgroup included measures that assessed conditions prevalent in a Medicaid population, including depression, schizophrenia, and substance use, in addition to measures that assessed utilization of general mental health services. The workgroup did not recommend including any of the five measures suggested in public comment as they concluded that these measures addressed similar content areas as other higher-rated measures or were rated very low in feasibility for State collection and reporting (e.g., HBIPS-4 Patients Discharged on Multiple Antipsychotic Medications, which requires a State to collect multiple hospital-level data elements).

Subcommittee Voting Results

The Subcommittee engaged in two rounds of anonymous voting: one round after workgroup deliberations and another round after a full Subcommittee review of the round one results.

In round one, 35 measures received a majority (greater than 50 percent) of "yes" votes to the question of whether it should be included in the core set. After this round, the full Subcommittee convened to review the resulting set as a whole. The Subcommittee discussed how each measure represented the conditions and populations relevant to Medicaid. In addition, the Subcommittee examined each measure's data source and whether it was in use in existing programs. In this way, the Subcommittee aimed to ensure measures accounted for issues important and relevant to Medicaid while keeping the core set from becoming too burdensome for States to report.

After this discussion, the Subcommittee participated in a second round of voting. In round two, 23 measures received greater than 50 percent "yes" votes to the question of whether it should be included in the core set. One measure, HIV/AIDS Screening: Members at High Risk of HIV/AIDS, received exactly 50 percent "yes" votes. A decision was made to explore the options for an HIV measure in the Initial Core Set with colleagues from the Centers for Disease Control and Prevention (CDC) and the Health Resources and Services Administration (HRSA).

The 24 measures that received at least 50 percent "yes" votes in round two were moved forward as the final measures being recommended by the Subcommittee for inclusion in the Initial Core Set.

Balancing and Finalizing the Set

After a final review of the resulting set of measures, CMS and AHRQ concluded it was important to include two additional measures in the final core set: PC-01 Elective Delivery and Timely Transmission of Transition Record. The elective delivery measure assesses patients with elective vaginal deliveries or elective cesarean sections at ≥37 and <39 weeks of gestation completed. This measure was recommended by the Maternal /Reproductive Health Workgroup, but was not prioritized over other measures during the Subcommittee's final vote. Research suggests elective deliveries are associated with adverse outcomes. Elective inductions, compared to spontaneous labor, result in more cesarean deliveries and longer maternal length of stay (3). Further elective deliveries before 39 weeks' gestation also result in higher rates of adverse respiratory outcomes, mechanical ventilation, sepsis, and hypoglycemia for newborns (4). As such, CMS and AHRQ concluded the measure was important to keep in the Initial Core Set.

The second measure that was added back into the Core Set is a transition record measure which assesses the percentage of patients discharged from an inpatient facility to home or another site of care for whom a transition record was transmitted to the facility or relevant health care professional within 24 hours of discharge. Much like the elective delivery measure referenced above, the care transition measure made it to the preliminary recommendations, but not through to final voting. The Workgroup found the measure important, but was concerned it may be difficult for some States to collect. CMS and AHRQ concluded that being able to measure care transitions is critical to the Medicaid population. While standards for proper discharges dictate that a transition plan be prepared for each patient and relayed in a timely manner to the clinical caregiver accepting responsibility for post discharge (5), a literature review found that direct communication between hospital physicians and primary care physicians occurred infrequently and that the availability of a discharge summary was low (6).

Additionally, CMS met with colleagues from CDC and HRSA to discuss the measure addressing HIV/AIDS. Half of the Subcommittee had voted to recommend the HIV/AIDS Screening: Members at High Risk of HIV/AIDS measure for inclusion in the Initial Core Set, but the group as a whole agreed to defer the recommendation to a subset of SNAC members after consultation with CDC and HRSA. Further discussion of the measure identified limitations to the scope of this measure. For example, screening for HIV is recommended for high-risk populations. However, the eligible population for screening in this measure is defined based on Medicaid claims data demonstrating that a person was screened, diagnosed, or treated for a sexually transmitted disease other than HIV/AIDS or screened for hepatitis B or C. Concerns were raised that this denominator would capture only a fraction of the population at risk for HIV/AIDS. In addition, Medicaid claims data would not capture screening performed in settings that do not bill Medicaid (e.g., screening provided by a free clinic). The decision was made to replace the screening measure with the HIV/AIDS: Medical Visits measure from the original draft list of measures, which assesses access to care for individuals with HIV/AIDS, as a more feasible way to address a condition critical to the Medicaid population. 

Thus, a total of 26 measures are being recommended for the final Initial Core Set (Table 1).

7. Initial Core Set of Children's Health Care Quality Measures.

Page last reviewed October 2014
Page originally created December 2010
Internet Citation: 4. Results. Content last reviewed October 2014. Agency for Healthcare Research and Quality, Rockville, MD.