Pick a small set of related health, patient experience, and cost measures to start

Pick a small set of related health, patient experience, and cost measures to start

Your measurement strategy can start with a very small number of health, patient experience, and cost measures that rely on existing data or that are feasible to collect. The measures should be related by topic or population. Data quality may be a work in progress. But, by getting started, you can then refine your measures and data with an iterative approach.


Earlier we discussed why you should measure the Triple Aim (step 1) and how to select a topic or population (step 2) to measure and analyze. The next step is to choose at least three measures (a health measure, a patient experience measure, and a cost measure) that are related to each other and to your chosen topic. We suggest choosing a small set of measures, maybe even only three measures, to focus your measurement project and to simplify interpretation.

It’s important to note that a single measure about health, patient experience, or cost does not stand in for the broad concept. For example, you may choose a measure about hypertension control as your health measure, but the single measure does not equate health for all patients in your community health center. Indeed, you might choose multiple measures for health, patient experience, and cost to reflect the wider population of patients at your community health center. Throughout the Toolkit, for simplicity’s sake, we use “health measure”, “patient experience measure”, and “cost measure” as flexible terms that could encompass a range of potential measures.

Picking measures can be overwhelming given the number of measures you’re already collecting and the hundreds of additional measures you could be collecting. Furthermore, it is critical to account for the time needed to get and process the data for these measures. Based on our work with community health centers, we suggest that quality improvement and analytic teams consider the following when choosing measures.

Include health measures that you already collect

You can develop a measurement strategy around existing measures you contribute to the UDS, collect to demonstrate meaningful use, or report for pay-for-performance programs. For example, you may decide to use an existing UDS measure as your health measure. Then you only need to identify related patient experience and cost measures to complete your measurement strategy. 

You can also look beyond existing reporting requirements for potential measures. In this video, Dr. Boutwell offers advice about how to use existing data to inform your measurement strategy.

Include patient experience measures

Patient experience is important to include because better experience is associated with less healthcare utilization, better adherence to recommended treatments, and better clinical outcomes. But patient experience can be difficult to measure because you must get data directly from the patients, rather than from more readily available administrative or clinical data.

Some community health centers use Consumer Assessment of Healthcare Providers and Systems (CAHPS) surveys to measure patient experience. CAHPS surveys are validated and feature readily available benchmarks that community health centers can use to compare their performance locally and nationally.

If you can’t administer a CAHPS survey, you may want to develop your own patient experience survey. To help, we suggest 27 patient experience survey items, organized by topic. Many of these items come from the Clinician and Group CAHPS surveys and have been cognitively tested to ensure that users understand what is being asked and that the questions capture needed information.

Community health centers also may consider other topics (e.g., patient and family engagement, patient activation) or measures from inventories described below. While your own survey probably won’t allow for benchmarking with other organizations, it can provide helpful information about patient experience.

Fill gaps in patient experience data by using proxy measures

You might consider using proxy access measures for patient experience. Examples include: check-in time to time to seeing a provider, time to next available appointment, number of dropped phone calls, and cycle times. These proxy access measures may correlate with patient experience and be easier to collect because they are from your own systems. Although useful, proxy patient experience measures provide limited information because you’re not hearing directly from patients.

Include cost measures

What constitutes a “cost” measure varies depending on who you ask— community health centers and providers, payers and health plans, or patients. Community health centers and providers consider costs as the dollar amounts to provide care for patients. From the payer and health plan perspective, costs are the actual amounts paid to providers for caring for patients, which often differ from the amounts providers charge. For patients, costs typically refer to out-of-pocket spending at the point of care.

Although challenging, it is important to include cost in your measurement strategy. Cost is an essential part of the Triple Aim, and cost is critical in understanding value. By including a cost measure now, even an imperfect one, you will be able to improve your cost data over time.

The Robert Wood Johnson Foundation’s Counting Change-Measuring Health Care Prices, Costs, and Spending and the Agency for Healthcare Research and Quality’s (AHRQ) Selecting Quality and Resource Use Measures: A Decision Guide for Community Health Collaboratives provide you with background, rationale, and short lists of cost measures.

Use total cost of care as your cost measure

Total cost of care accounts for the costs of all services provided to patients, including professional, inpatient and outpatient, pharmacy, lab, radiology, ancillary, behavioral health, and other services. Total cost of care is risk-adjusted and reported as a per member, per month amount. Total cost of care data typically rely on health plan administrative claims.

In this video, Dr. Boutwell explains why to measure the total cost of care as part of the Triple Aim. 

Challenges with total cost of care include:

  • Getting timely data from health plans. Available health plan data may be several months or even years old.
  • Requesting and processing data. You may need to build in considerable time to link health plan and community health center data as patient identifiers are often different.
  • Compiling data from multiple health plans. You likely work with multiple health plans or participate in different payment incentive programs. Total cost of care data from a given health plan only represents costs for a subset of patients you see.
  • Analyzing cost for specific patient groups. Total cost of care data for all members of a health plan may not apply just to a specific set of patients—for example, only patients in your hypertension quality improvement program. 

Despite these challenges, total cost of care is a complete measure that is standardized, relatively stable, and measurable over time. As a result, even if you currently don’t have access to total cost of care data, moving toward a total cost of care measure is critical, especially in the shift from fee-for-service to value-based payment.

See AHRQ’s Measure Summary for more detailed explanations of total cost of care.

Fill gaps in cost data by using proxy measures

When total cost of care data is unavailable—for example, you can’t access payer claims data—you can include a proxy measure in your Triple Aim measurement strategy. Proxy cost indicators typically measure resource utilization, such as 30-day readmissions, ambulatory sensitive inpatient hospital admissions, and non-emergent use of the emergency department. The rationale is that reducing unnecessary resource use will lead to reduced costs.

Even if your cost measure or data are imperfect, any progress toward measuring cost will get you on the path of demonstrating value.

Review measure inventories for ideas about measures

You may find that existing measures aren’t sufficient for your measurement strategy. This list of measure inventories can help you identify new measures to pursue. Looking through measure inventories allows you to explore the range of measures that align with your chosen topic or population.

Please be aware when reviewing measure inventories that many measures have only subtle differences. Be sure to choose measures that fit your topic or your population, and consider appropriate time horizons for measurement. For instance, long-term chronic disease outcomes may be more difficult to capture than shorter-term measures.

Prioritize to narrow the measure set

A structured selection process can help you narrow your choices if you are deciding among many promising health, patient experience, and cost measures. You don’t have the resources to study all measures, so how do you pick what to focus on? This resource from Buying Value called “How to Build a Measure Set” helps users articulate measurement goals and to develop and apply measure selection criteria to narrow the list to the best measure candidates and finalize measures. 

Case study part 2: Hypertension quality improvement program

The following is a continuation of the case study begun in Step 2. Next, Part 2 of the case study will demonstrate how to identify and choose a health, patient experience, and cost measure that will help you analyze value.

As the quality improvement officer at your community health center, you chose to focus on your new hypertension quality improvement program (step 2) to kick off your Triple Aim measurement project. To analyze the value of the program, your next step is to choose a health measure, a patient experience measure, and a cost measure that all pertain to patients in this program at all nine clinic sites within your organization.

  • First, for potential health measures, you review your existing measures and find that you are already reporting several different blood pressure control measures to the Uniform Data System and payers. You’re even able to pull two of the measures for the specific patients in your hypertension quality improvement program – percent of patients with hypertension with blood pressure levels under control (<140/90 for patients aged 18-59 or <150/90 for patients aged 60-85) and adherence to prescribed antihypertensive medications.
  • Next, you review your existing patient experience measures to determine if any are influenced by taking part in the hypertension quality improvement program. You think back to your annual patient experience survey with its mix of validated items. One item in particular seems relevant: “When I left the clinic, I clearly understood the purpose for taking each of my medications.” The patient experience survey also has an average satisfaction score. But since the existing survey doesn’t include patient identifiers, you cannot pull survey responses from only the patients in the hypertension quality improvement program. Instead, you think about asking a simple question to patients in the hypertension quality improvement program in order to gauge patient satisfaction, and make a note to yourself to discuss possible measures with staff.
  • Finally, you think about what measures to use for cost. You decide to approach your Medi-Cal health plan to get claims data. But until you form a data partnership with the health plan, you must consider other cost measures in the interim. You can look at your community health center’s costs of delivering care to hypertension quality improvement program participants, but can also consider a resource utilization measure that you are already gathering: the ambulatory sensitive hospital admission rate. Here, you are able to pull data on just the hypertension quality improvement program participants at all of the nine sites in your health center organization. That might be simpler than computing the health centers’ costs of delivering care to the quality improvement program participants.
  • Now, you have a few measures to choose from, so you ask yourself some questions. Can you get the data you need on a reliable basis? Are the data reasonably accurate? How difficult are the data to obtain?

For health, you choose controlled hypertension. You are already reporting this data out to UDS. Since you have a gap in patient experience data, you decide that you need to discuss a new survey with your clinical staff. For cost, you decide to use the ambulatory sensitive hospital admission rate among program participants. Since your health center already uses this Agency for Healthcare Research and Quality (AHRQ) measure, it won’t pose an additional measurement burden to gather this data for participants in your hypertension quality improvement program. To think ahead about improving data for the future, you also decide to contact your health plans to begin to gather data on total cost of care for all the clinic sites in your health center organization.

An example of working to reduce avoidable readmissions

Dr. Boutwell explains in this video how to develop a measurement strategy that includes health, patient experience, and cost measures for the example topic of avoidable readmissions. Although the video illustrates how to develop a measurement strategy around a single topic, the thought process can be applied to any topic, not just avoidable readmissions.

See what your colleagues are doing

Here is a Triple Aim measurement strategy from other community health centers.

Topic or population

Health measure

Patient experience measure

Cost measure

Complex care case management for clinic patients with highest emergency department and hospital utilization

Number of behavioral health follow ups for patients in complex care case management program

Percent of patients in complex care case management program who were satisfied with program

Emergency department visit rate


Hospital readmission rate

An additional resource to identify priority clinical areas, along with descriptions of associated high-value care and measures, can be found starting on page 15 of KPMG International’s report, Measuring the Value of Healthcare delivery: Cutting through complexity.