Feasibility Testing of the Quality-monitoring Tool, Qdact, for the Palliative Care Research Cooperative

Healthcare processes are measured, evaluated, and characterized through the use of healthcare quality measures. Healthcare quality measures are tools that quantify the consistency of care delivery within a population eligible for the process. Containing a numerator (those with successful delivery of a care process) and a denominator (eligible patients for that care process), quality measures produce a frequency or adherence rate to which a care process was performed. Adherence rates can then be compared to evaluate quality of care across clinicians, organizations, or collaborations to compare data, establish benchmarks, and spur quality improvement projects.

In palliative care, for example, it is usually considered best practice to prescribe opioids for moderate/severe pain. Imagine that a palliative care program's calculated adherence rate reveals that their clinicians prescribe opioids for moderate-to-severe pain only 50% of the time. Armed with this information, the program can now develop a directed quality improvement project with the intention of improving this performance towards a more ideal goal (e.g. 75%). Further, it can provide feedback to clinicians in real-time regarding how they are performing against the quality measures of interest. In this example, clinicians may receive an electronic alert reminding them to prescribe an opioid when directed by an accepted best practice. This real-time approach, combined with a system that promotes culture of data collection, sharing, benchmarking, and reporting, are effective methods to improve healthcare quality. Lastly, and the focus of this proposal, is to build and test such an infrastructure that performs such real-time quality monitoring of healthcare measures in the Palliative Care Research Cooperative group (PCRC).

The investigators have previously identified the three major components needed for an effective and usable quality-monitoring infrastructure. Together, these three components answer the "what", "how", and "for why" questions that must be addressed within a quality assessment and improvement system.

First, is the ability to perform collaborative and integrated data collection across several sites. Successful multi-site data collection requires a centrally governed set of data collection processes, which are guided by a data dictionary. A data dictionary is a set of agreed-upon data elements, answer choices, rules, and branching logic. The data dictionary informs the development of a data collection platform for use by clinicians. Together, the data dictionary and software for use by clinicians guide "what" data is collected, and ensure that the intended collaborative analyses can be performed with the data set created.

Second, is the process for data collection - the "how" characteristic within the system. Data is collected, transmitted and recorded through the use of a data collection platform, transmission processes, and registry, respectively. The data collection platform is the interface in which real-time data is captured and recorded. This can involve paper-based or electronic forms using patient, caregiver, or clinician reporters. Data is then transmitted to the registry, either through electronic or manual means. Lastly, data is collected and securely stored in a prospective registry, so quality reports can be generated and research analyses completed. These steps are recommended standards for development of health information technology by the Agency for Healthcare Research and Quality (AHRQ).

Third, is the component of the infrastructure that answers the question, "for why?" Several reports have highlighted the need to translate raw data from quality monitoring efforts into continuous feedback on quality to clinicians and other end-users to motivate the delivery of best practices. This allows for changes in clinician performance during usual clinical care delivery, thus meeting the Institute of Medicine's aim for a rapid learning healthcare system. Generally, feedback is provided through system-generated reports that target specific end-users (e.g. clinicians, administrators) delivered during pre-specified time periods (e.g. weekly, quarterly).

At Duke University, investigators recently built the information technology infrastructure needed for prospective quality measure adherence and outcomes monitoring in palliative care. This system was developed and deployed in the Carolinas Consortium for Palliative Care, a four-site collaboration between Duke University and three community palliative care organizations. Recently, this Consortium has expanded to include organizations outside the Carolinas; eleven sites now comprise the Global Palliative Care Quality Alliance (GPCQA). The rapid expansion of qdact users and subsequent data collected have supported several research-level analysis published in the literature.

In using qdact.pc, clinicians record data on processes of care and patient-reported outcomes on personal iPads® during face-to-face clinical encounters with patients. During patient interviews, clinicians record patient-reported areas of distress, clinical management decisions, and patient-reported outcomes using validated instruments. These instruments include those common to the field, including the Edmonton Symptom Assessment Scale (ESAS), Palliative Performance Scale, and FACT-G. Longitudinal changes in these scales are captured through repeat use of qdact.pc during subsequent encounters. Further, qdact.pc calculates length of stay from admission and discharge dates, changes in symptom severity by calculating the difference between two dates, and readmission rates by analyzing whether patients in the registry had previously been admitted. Then, care processes and outcomes can be linked using these data.