JOURNAL OF THE LOUISIANA STATE MEDICAL SOCIETY
SUCCESSFUL REDUCTION IN CONGESTIVE HEART FAILURE READMISSION RATES: A QUALITY IMPROVEMENT PROJECT N. Turaga, MD; S. Gavini, DO; H. Manthena, MD; J. Patel, MD; G. Jacob, MD; S. Yasin, MD; I. Shuja, MD; A. Narayana, MD; S. Saad, MD; A. Jakkoju, MD; E. Borrero, MD Department of Internal Medicine, LSU Health Sciences Center – Lafayette Introduction: Congestive heart failure (CHF) is a major healthcare issue all over the world with an increasing morbidity and economic burden. CHF has a high 30-day hospital readmission rate. However, detailed description of the causes and preventability of readmissions are lacking. We performed a quality improvement study to understand the high impact factors causing CHF readmissions and to formulate effective strategies to reduce readmissions. Methods: A multidisciplinary team consisting of residents, attendingphysicians, casemanagers andnurseswas formed, and the studywas done in four phases. In the first phase retrospective chart reviewof patients admitted for CHF between July 2016 and November 2016 was done. Patients readmitted within 30 days were identified and the potential causes of readmissions were analyzed. During the second phase, interventions to address the high impact factors were formulated and the most feasible ones were selected with a common consensus. Third phase was the implementation phase in which the interventions were implemented from August 2017. Fourth phase was analyzing the project implementation and results. PDCA methodology was used as a quality improvement tool for each phase to help identify factors and processes in our institution that required change. Results: Retrospective chart review in first phase revealed 1-month readmission rate of 22 % and 1-week readmission rate of 11%. The major patient related factors impacting readmissions were medication unaffordability (75.5%), medication noncompliance (28.5%), and transport (10%). While the major health care system related factors were lack of 1-week post discharge follow up (100%), medication reconciliation and patient education. Interventions selected for implementation were educating themultidisciplinary teamabout CHF guidelines and the interventions, establishing 1-week post discharge clinic follow up, $40 financial assistance for patients at discharge and providing a new detailed educational material to the patients. 3 months after the interventionsweremade, 1-month readmission dropped down to 10.1% and 1-week readmission rate dropped down to 0%. Conclusion: Every medical facility has its own unique patient population and hurdles in providing health care. An individualized approach to identify and address those hurdles can lead to successful strategies to reduce CHF readmissions.
CLINICAL PREDICTION TOOLS VS CLINICAL GESTALT: A DIAGNOSTIC DILEMMA
A. Domney, MD; M. Memari, MD; C. Chakraborti, MD Department of Internal Medicine, Tulane University Health Sciences Center, New Orleans
Case: A 53 year-old-man presented with one day of dyspnea, syncope, and chest pain that was partially relieved with nitroglycerin; however, a pleuritic component that was not. The patient’s past medical history included severe CAD, hypertension, and COPD. At admission, supplemental O2 was required to keep saturation 95-99%. Bilateral, diffuse, rattling breath sounds were appreciated, while the remainder of exam was benign. EKGs and cardiac enzymes remained normal. Chest radiography was furthermore unremarkable. Ultimately, CT imaging revealed pulmonary emboli (PE) in right lower, left lower, and left upper lobes. Discussion: Diagnosing PE remains a challenging proposition for the general internist. Utilizing Wells’ Criteria to estimate pretest probability of PE is common and often stratifies the risk of diagnosis based upon initial clinical suspicion using six objective and one subjective criteria. A previously reported retrospective analysis on suspected PE patients comparing the utility of gestalt assessment, Wells’, and the revised Geneva score revealed that gestalt assessment was overall superior (AUC 0.81) in selecting patients both with low and high probability of PE diagnosis compared to clinical decision rules (Wells’ AUC 0.71; revised Geneva AUC 0.66). In patients where clinical gestalt predicted high likelihood, PE prevalence was found to be ≥55%. Results from this study suggested that overruling clinical tools based upon physician intuition might actually improve the rules’ performance. The characteristic symptomatologies of a PE were not evident in our patient. In our patient, the clinical likelihood of a PE centered around the subjective criterion of the Wells’ score: “was PE the most or equally likely diagnosis?”. worth 3 points on the scale. Without fulfilling any objective criteria, the patient’s initial score (0) would stratify to unlikely PE and low risk. When including a positive answer to the subjective criterion, a Wells’ score of 3 results in an unlikely, but moderate risk group, highlighting the value of clinical gestalt.
J La State Med Soc VOL 170 MARCH/APRIL 2018 67
16
Made with FlippingBook Digital Publishing Software