Models to optimise medication safety in elderly and oncology inpatients

Šarčević, Ana (2018): Models to optimise medication safety in elderly and oncology inpatients. Dissertation, LMU München: Medizinische Fakultät

Abstract

Medication therapy is recognised as a high-risk process. German Federal Ministry of Health has proposed certain measures to ensure an optimal medication process with the aim of reducing medication errors and thereby avoidable risks for the patient in drug therapy. The concept is called medication safety. Its centre of attention is fostering awareness of medication safety among patients, physicians, pharmacists and caregivers, to improve information on medicines and intersectional communication about drug therapy as well as facilitation of medication safety research. A central point represents the need for greater involvement and information of the patient with regard to his medication therapy. The focus of this thesis is to propose models to optimise medication safety in the hospital setting. The first part of the thesis presents a study protocol for multi-professional model to reduce drug-related readmissions in care-dependent elderly. The protocol is based on the foundation of Spirit 2013 statement, which defines standard protocol items for clinical trials. The model recognises the need for safer medication use through the concerted efforts of all caregivers. It has been developed and recently tested in a pilot study. First results showed the model’s tendency toward active collaboration among the ward-based caregivers and outlined a positive impact on medication safety. However, its effectiveness has still to be assessed in a full-scale evaluation with an adequate sample size. The planed evaluation study is an open randomised controlled trial supplemented by a qualitative evaluation of the intervention. During the hospital stay, patients are randomly allocated to either a control or an intervention group. Control group represent current medical standard of care where the intervention represents multi-professional care according to the defined Standard Operating Procedure for this purpose. After the hospital stay, both patient groups are followed up in the same manner for 12 months to assess the long-term efficacy of the model with a patient-relevant outcome – drug-related hospital readmissions. Qualitative assessment in this evaluation study consists of patient interviews at the end of the hospital stay and focus group for care providers after the study intervention time. Integrating qualitative methodologies implies studying individuals in their natural setting to characterise and provide potential explanations for particular events and behaviours. Evaluation of the model using mixed-methods approach should provide a useful example for further improvements of multi-professional complex intervention models and medication safety in Germany. The second part of this thesis presents the model of optimised pharmaceutical care service on the oncology ward and development of a prediction model. It is a pilot study outlined by STROBE and TRIPOD statement. Pharmaceutical care service was assessed in terms of management of drug-related problems (DRPs). The aim was to develop DRP prognostic model. The pharmaceutical care model was further assessed in terms of number, type and risk factors of DRPs leading to intervention, type of intervention provided, and PRO-CTCAE symptom burden and impact on pharmaceutical intervention. The most important measurements were: German PRO-CTCAE core item set (weekly assessment, recall period 7 days, 12 symptom item clusters, three dimensions: frequency, severity, interference), adverse drug reaction risk score on admission, APS-Doc® classification system for drug-related problems and Doku-PIK documentation system for pharmaceutical intervention, both developed for the hospital setting. In the study sample of 101 patients, 46 were women. Average age was 65 years and average stay on the ward was 10 days. The most frequent diagnoses were malignant neoplasm of bronchus or lung, diffuse large B-cell lymphoma and multiple myeloma. On average each patient had two DRPs leading to pharmaceutical intervention. The most common drug-related problems were drug-drug interactions, drug dosage, drug prescription/monitoring, indication and contraindication. The most frequent pharmaceutical interventions were symptom surveillance, information to physicians and nurses, initiated diagnostic tests, stop/pause the drug and change of the drug dose. The most frequently reported PRO-CTCAE symptoms were fatigue, anxiety and sadness, and pain. For 13 patients pharmaceutical recommendations were based on the PRO-CTCAE items. The implementation rate of pharmaceutical recommendations by the ward-based team was 93 %. Poisson regression model was performed to assess risk factors predicting pattern of DRPs leading to pharmaceutical intervention on the oncology ward. In the second step, the regression was used to combine multiple predictors by assigning relative weights to each predictor to obtain a probability of DRP during the stay on the oncology ward. There are two models to be distinguished: the initial or up to 5th day on the study ward, and follow-up model from the 5th day up to 10th day on the study ward. Significant risk factors in the initial model were ECOG performance score, adverse drug reaction risk score and presence of heart failure; in the follow-up model ECOG performance score and presence of renal failure. Prognostic ability of both regression models was assessed with ROC curve showing AUC of 0.777 (95 % CI 0.686 – 0.868, SE 0.046, p < 0.001) and AUC of 0.683 (95 % CI 0.532 – 0.835, SE 0.077, p = 0.033), respectively. Based on the initial regression model, DRP prognostic model has been proposed (AUC of 0.790 (95 % CI 0.697 – 0.883, SE 0.048, p < 0.001), with a scoring system from 0 to 10 and cut off at 5. That means patients with 5 or more points are according to the Initial DRP prognostic model highly probable to experience DRP leading to a pharmaceutical intervention during the five-day stay on oncology ward. The pilot study has showed that cancer patients exhibit many DRPs requiring a pharmaceutical intervention. PRO-CTCAE may support pharmaceutical care on the oncology ward. The high acceptance of the intervention indicates the need of a pharmacist as integrated part of the oncology ward-based team. The initial DRP prognostic model may assist pharmacist to prioritise their service and to optimise their workload. But small study sample and lack of internal validation limits score interpretation, considering current results provisional and calling for the score development and validation in a study with bigger sample size. The approach, however, provides the methodology for further research of the most optimised patient-oriented, quality-driven, outcome-based oncology ward teamwork.

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Keywords: medication safety, complex intervention, mixed-method approach, pharmaceutical care, elderly, oncology, risk model
• Themengebiete: 600 Technik, Medizin, angewandte Wissenschaften ; 600 Technik, Medizin, angewandte Wissenschaften > 610 Medizin und Gesundheit
• Fakultäten: Medizinische Fakultät
• Sprache der Hochschulschrift: Englisch
• Datum der mündlichen Prüfung: 21. Dezember 2018
• 1. Bericht­erstatter:in: Mansmann, Ulrich