������Ƶ

Potential Measure Development Methodology

The potential measures were developed through an iterative process that included review of the CCRM Atlas, generation of ideas for potential measures, and periodic review and discussions with the CCRM Expert Panel.  After the full set of potential measures was developed, staff reviewed both the potential measures and the existing measures in the CCRM Atlas to identify a core set of 13 measures to prioritize for future development. The core set reflects a judgment about which measures may be the most useful and feasible for quality improvement and program evaluation. The core set of measures begins at /prevention/resources/chronic-care/ccrm-atlas-suppl/ccrm-atlas4.html.

2.1 General Measure Attributes

In general, measures intended to be used for quality improvement or program evaluation should meet several basic criteria:

• First, measures must be scientifically sound. In the case of CCRM, this means that the focus should be on evidence-based clinical preventive services suitable for delivery through clinical-community relationships. Applicable services were selected early in the project and are described in the CCRM Atlas. Further, measures of the structures, processes, or outcomes associated with the delivery of these services through a clinical-community relationship should be based on established conceptual models and use validated data collection instruments where applicable. Without exception, the potential measures presented in this report need additional development work to establish their scientific soundness. As discussed in the CCRM Evaluation Roadmap, there are key unanswered questions related to the conceptual models underlying the design and implementation of clinical-community relationships themselves that must be resolved prior to or in conjunction with the measure development.  Even where the potential measures draw on existing measures used in other situations, these measures need to be validated in the context of clinical-community relationships.
• Second, measures must be relevant to the clinicians, community-based organizations, or patients involved in clinical-community relationships. For clinicians and service-delivery organizations, the measures must assess important aspects of service delivery for which they view themselves as accountable, and which they have the potential to improve. For patients, measures should be relevant to the decision to seek and complete preventive services or to the choice of service providers or settings.
• Finally, the measures must be operationally feasible. This includes the availability of needed data and sufficiently large denominators for reliable assessment. Possible data sources for each potential measure are indicated in the measure descriptions.

2.2 Potential Measure Context and Limitations

This supplement provides as much detail as possible in the potential measure definitions, but these definitions should be viewed as concepts for future measure development, testing, and adaptation rather than final specifications. Through the process of developing the list of potential measures, we identified several complexities and limitations that provide important context for understanding the measure set and how it might be used.

1. Potential Measures Organized According to the CCRM Measurement Framework – This set of potential measures is organized according to the measurement framework described in the CCRM Atlas. The framework focuses on the structures, processes, and outcomes associated with the three core elements of clinical-community relationships (clinics/clinicians, community resources, and patients) and the relationships among these elements. The potential measure set does not include measures that are external to the domains within the framework, although there are broader delivery system and community factors beyond the framework that may be relevant to measure in some situations. Instead, this supplement focuses on filling in gaps in the collection of existing measures relative to the measurement framework.
2. Lack of Programmatic Context – In many cases it proved difficult to develop detailed potential measure specifications in the absence of information about the specific program being assessed or the anticipated uses of the measurement data. Adding to the challenge, for most of the measures there is no external referent or standard describing a recommended process or approach to which the measure can be tied in the way that clinical quality measures can be tied to an evidence-based practice guideline. These potential measures should therefore be viewed as a starting point for future measure development and refinement work in the context of specific program implementation efforts, quality improvement initiatives, or program evaluations.
3. Unit of Measure Reporting – To help clarify the intent and description of the potential measures, the investigators drew on insights from the Institute of Medicine’s Primary Care and Public Health Framework for Action¹ which discussed the traditional focus of clinicians on providing medical services to individual patients with immediate health needs, while public health focuses on offering a broader array of ser­vices across communities and populations. This highlights the need to be clear about the unit of measurement for each potential measure, although this is often ambiguous because of the absence of context on the use of the measure. Many of these measures could be reported at different levels depending on the measurement purpose. For example, measure OO – strength of a clinical-community relationship-- could be reported at the level of a particular clinical-community relationship between two organizations, at the level of an organization reflecting all of the clinical-community relationships in which it is engaged, or at the level of the community as a whole reflecting all of the relationships between organizations within the community. Similarly, many patient-focused measures could be reported for patients of a particular clinician/clinic, patients of a particular community resource, or all patients in the community as a whole. In describing the calculation methods for the measures we have tried to indicate the reporting levels that might be useful, but the measures may also be useful at other reporting levels not mentioned.
4. Definition of “Community” – The concept of “community” is fundamental to the discussion of clinical-community relationships and can have multiple interpretations. This document recognizes potential ambiguities in the use of the term community and acknowledges the need for additional research to define it appropriately for use in clinical-community relationship implementation and measurement. One might take a geographic or demographic approach to defining communities based on census classifications. However, the CCRM Expert Panel thought that a broader definition that comes from a public health perspective, in which a community is defined as a “group of people with diverse characteristics who are linked by social ties, share common perspectives, and engage in joint action in geographical locations or settings”² might be more useful. The Panel also acknowledged that modern telecommunications may be reducing the importance of geographic boundaries in the formation of communities, adding to the complexity of defining units of measurement appropriately.
5. Definition of “Patient” and “Client” – The concepts of patient and client are also fundamental to the discussion of clinical-community relationships.  The patient is one of three elements that form the base of the conceptual framework and refers to the individual who receives primary care services, including preventive care and chronic illness care, delivered in a clinical setting. If a measure refers to the delivery of services in a community setting or the relationship between an individual and community resource, the term client is used to refer to the individual receiving services, reflecting standard usage in a social service context. The terms patient and client are used in the current context of primary health care reform as described by multiple patient-centered medical home initiatives occurring nationally. 
6. Data Sources – Each potential measure description includes at least one potential data source. These data sources are suggestions, but other sources of data may turn out to be more useful or feasible. For measures of complex concepts that may be collected through survey instruments, the measure will not necessarily be based on a single survey question. Future development work can help define valid and reliable questions that can be combined to form a coherent composite measure. For measures with a suggested data source of an organizational audit, further development work would focus on testing assessment forms and designing reliable processes for gathering needed information to support measurement.
7. Broader Applicability of Measures Beyond Clinical-Community Relationships for Prevention – While the potential measures have been defined in the context of clinical-community relationships for the delivery of selected clinical preventive services, many of the measures could also be useful applied in the context of the delivery of other types of services (e.g., chronic disease management services) through a clinical-community relationship. Further, some of the measures may be applicable to the delivery of services not involving a clinical-community relationship at all.

Because this supplement describes measure concepts rather than detailed specifications, future measure development and adaptation work will need to take into account the availability of data and the specific uses of the measures for performance improvement, evaluation, or research.

2.3 Potential Measures Template

Each potential measure is presented in a standard template in the following format:

^1. Clinical-Community Relationships Measures (CCRM) Atlas. March 2013. ������Ƶ, Rockville, MD. http://www.ahrq.gov/prevention/resources/chronic-care/clinical-community-relationships-measures-atlas/index.html

^2. Clinical-Community Relationships Evaluation Roadmap. July 2013. ������Ƶ, Rockville, MD. http://www.ahrq.gov/prevention/resources/chronic-care/clinical-community-relationships-eval-roadmap/index.html

^3. Primary Care and Public Health: Exploring Integration to Improve Population Health. March 2012. Institute of Medicine.

^4. MacQueen, K.M., McLellan, E., Metzger, D.S., Kegeles, S., Strauss, R.P., Scotti, R., Blanchard, L., Trotter, T.T., II. What Is Community? An Evidence-Based Definition for Participatory Public Health. Am J Public Health: December 2001, Vol. 91, No. 12, pp. 1929-1938.