Improving Outcomes by Measuring and Analyzing Data

National Harbor, Maryland—With advances in technology, physicians are constantly gaining access to more data about their patients. However, many providers have not embraced the idea of tracking and using the information. During a session at the AAPM meeting titled Using Outcomes Data to Improve Patient Care, healthcare professionals emphasized that utilizing the data in an effective manner can help decrease errors, cut costs, and lead to medical advances. Michael Ashburn, MD, MPH, MBA, director of pain medicine and palliative care at the University of Pennsylvania, Philadelphia, began the session by saying that physicians have numerous options when treating patients, although they are not immune to making mistakes. Citing commentary from the Journal of the American Medical Association [2010; 304(2):204-205], Dr. Ashburn said that an estimated 100,000 patients die of healthcare-associated infections, between 44,000 and 98,000 die of other preventable errors, and an additional ≥10,000 die of diagnostic errors or not receiving recommended therapies. Dr. Ashburn said errors are commonly attributed to teamwork issues such as a lack of communication or a failure to heed the advice of others. Another problem is that most medical trials are uncontrolled, only involve a small number of patients, and include anecdotal data. To improve patient care, Dr. Ashburn suggested reducing variation through clinical pathways and standardized care plans, comparing quality measures with other healthcare organizations, and supporting innovation, risk-taking, and testing of new models. He emphasized basing physicians’ compensation on outcomes and sharing the outcomes data with physicians, who will be able to see how they compare with other physicians. Gilbert J. Fanciullo, MD, MS, director of the pain management center at the Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire, then discussed the Pain Outcomes Evaluation Tool (POET), a clinical database that collects outcomes such as a pain assessment, quality of life, function, and opioid risk screening. Dr. Fanciullo said the database includes approximately 40,000 patient visits and 15,000 patients who were treated at the Dartmouth-Hitchcock Medical Center for the past 3 years. Patients who visited the hospital for pain-related reasons were asked to complete surveys during their initial visit and during a follow-up period, as well. The questions included the type of pain medication used, the frequency of the pain medication taken, and the effect the pain medication had on the patients’ symptoms. The researchers also inquired about the patients’ initial expectations regarding how the pain medication would work, how the pain medication actually worked, and any significant physical or psychosocial events that occurred. Dr. Fanciullo said that POET allows providers to track their patients’ progress and is useful for clinical care, research, quality improvement, and reporting purposes. Marlin D. Cheatle, PhD, director of behavioral medicine at the University of Pennsylvania’s Pain Medicine Center, followed by emphasizing that pain management physicians must disseminate information to primary care physicians (PCPs) to collaborate in caring for patients and tracking potential opioid abuse. Dr. Cheatle cited the 2008 National Survey on Drug Use and Health from the Substance Abuse and Mental Health Services Administration that showed during 2007, of those who abused drugs, >2.2 million people ≥12 years of age took pain relievers; that number was greater than those who used any illicit drug. Studies have indicated between 3% and 62% of patients with chronic noncancer pain exhibited problematic opioid-taking behavior, according to Dr. Cheatle. Several risk factors are associated with patients taking opioids developing an addiction, including a personal or family history of substance abuse, age, history of preadolescent sexual abuse, mental abuse, and cigarette dependency. According to a risk stratification model that Dr. Cheatle discussed, patients can be categorized as low risk, moderate risk, or high risk for abusing opioids. The low-risk group includes patients who have no personal or family history with substance use disorder (SUD) or no or minimal comborbid psychopathology. They are recommended to see PCPs. Patients in the moderate-risk group meet with PCPs and specialists and include those with a personal or family history of SUD or a psychiatric disorder. The high-risk group patients undergo specialty pain management and include patients who are addicted to opioids or have a major psychiatric disorder. To assess the risks of opioid therapy, clinicians use screening tools, urine drug screening, medical record audit, and psychological examinations. Dr. Cheatle said guidelines proposed in 2009 recommend that patients with noncancer chronic pain who are taking opioids should have periodic urine drug screenings. He warned, though, that a positive test does not necessarily indicate a diagnosis of drug addiction, physical dependence, or impairment, and that an absence of a prescription opioid in a urine test could mean the patient may be hoarding the opioids. According to Dr. Cheatle, depression and anxiety are commonly found in chronic noncancer pain patients. In addition, major depression is found in 35% to 85% of cancer patients with pain, whereas anxiety disorders are found in 7% to 60% of cancer patients with pain. Also, an estimated 10% to 50% of patients on pain treatments suffer from posttraumatic stress disorder. Implementing a patient management system, using risk tools, and analyzing clinical outcomes information are all ways that healthcare professionals can better understand and serve patients with chronic pain, according to Fred N. Davis, MD, principal of ProCare Research, LLC, Grand Rapids, Michigan. Dr. Davis discussed the company’s PRISM systemthat measures patients’ pain level and includes the Pain Assessment Matrix (PAM) and the Pain Health Assessment (PHA). PAM is a scoring system comparing characteristics common in patients with chronic pain. PHA is a multidimensional survey in which patients report the effects of pain; the survey takes into consideration traditional pain scales, functional impairment assessment of activities of daily living, and social impairment. Patients take an initial PHA that evaluates their pain state and then take cumulative PHAs at 3 months, 6 months, and every 6 months thereafter to determine how their health has changed over time. The PRISM system includes a graphical display of results from the PHA during the initial survey and subsequent periods and reflects functional, psychosocial, quality-of-life, overall pain rating, and patient satisfaction measurements. Dr. Davis said data derived from PHA is useful in tracking clinical outcomes and as part of a care management system allowing physicians to use the data to make more informed decisions. He also suggested that the data could help clinicians in the future by improving research methods and clinical practices and supporting value-based care.