Pain Assessment & Intervention Needs (PAIN)

描述（由申请人提供）：本申请将使用二级数据分析来检查学术医疗中心的急性术后疼痛管理实践，使用从电子健康记录（EHR）和药物管理系统创建的大型数据集。本研究的主要目的是通过假设检验和数据挖掘来分析疼痛管理。目前的疼痛管理实践是由美国疼痛协会传播的标准，反映了如何治疗疼痛的最佳循证知识，但实施仍然缓慢。疼痛仍然是一个重要的术后问题，导致患者焦虑的程度可能超过对手术结果不成功的恐惧。改善疼痛管理的挑战之一是通过文件审计对护士行为（即评估、干预和重新评估）进行监督。检查电子护理文件可能提供发现知识和改进实践方法的机会。在这项研究中，对疼痛治疗指南的依从性将在电子病历中进行测量，主要通过患者护理的护理文件，包括药物管理和护理流程表。范德比尔特大学医学中心作为临床应用信息系统（如条形码药物管理图表和护理文件流程图图表）实施的先驱，可以在所有护理接受者的大量人群中测量基于证据的标准的实施水平，而不考虑年龄、性别或种族。我们计划使用数据分析来估计住院患者急性疼痛的患病率、严重程度、治疗和治疗的有效性。缓解疼痛将是主要的结果。我们将分析疼痛评估、干预和重新评估与疼痛评分严重程度的时间相关性，这些与计划（与需要）药物、多模式药物和行为干预有关。本研究将使用回归分析来检验实施疼痛标准对患者住院期间疼痛程度的影响。疼痛评分（在0-10的标准范围内）将通过重新评估的时间、重新评估的次数、阿片类药物给药、非阿片类镇痛药给药、非药物治疗和同时给药的治疗次数（干预强度）来预测。此外，数据挖掘将执行提取隐式的，以前未知的，有用的信息，以加强疼痛管理。本研究的长期目标是确定使用这些电子数据进行持续监测的可行性，这些监测可以通过可能的增量干预来提高绩效，类似于一些卫生保健机构实施的护士领导的快速反应小组的急性疼痛服务扩展。

DESCRIPTION (provided by applicant): This application will use secondary data analysis to examine acute post surgical pain management practice in an academic medical center using large datasets created from electronic health records (EHR) and medication administration systems. The primary aim of the study is to analyze pain management by both hypothesis tests & data mining. Current pain management practice is informed with standards disseminated by the American Pain Society reflecting the best evidence-based knowledge of how to treat pain, yet implementation remains slow. Pain remains an important post operative issue, resulting in a level of patient anxiety that may surpass fear of a unsuccessful surgical outcome. One of the challenges in improving pain management is the surveillance of nurse behaviors (i.e., assessment, intervention, and re-assessment) through documentation audits. Examination of electronic nursing documentation may present opportunities to discover knowledge and ways to improve practice. In this study, adherence to pain treatment guidelines will be measured in the electronic medical record, primarily via nursing documentation of patient care, including medication administration and nursing flow sheets. Vanderbilt University Medical Center, as a pioneer in implementation of clinical application information systems, e.g. bar code medication administration charting and nursing documentation flow sheet charting, can measure the implementation levels of evidence based standards in large populations of all recipients of care regardless of age, gender or race. We plan to use the data analysis to estimate the prevalence, severity, treatment, and effectiveness of therapies to manage acute pain in hospitalized patients. Pain relief will be the primary outcome. We will analyze the temporal association of pain assessment, intervention, and reassessment to severity of pain scores in relation to scheduled (versus as needed) medications, multimodal medications, and behavioral interventions. The study will use a regression analysis to examine the outcome of implemented pain standards on the level of pain experienced by the patient during hospitalization. Pain scores (on the standard scale of 0-10) will be predicted by time to reassessment, number of reassessments, opioid administration, non-opioid analgesic administration, non-pharmacological therapies, and the number of therapies administered together (intensity of interventions). In addition, data mining will be performed to extract implicit, previously unknown, and useful information for enhancing pain management. The long term goal of this study is to determine feasibility of using this electronic data for ongoing surveillance that could be monitored to improve performance using a possible incremental intervention similar the nurse led acute pain service extension of rapid response teams implemented by some health care institutions. PUBLIC HEALTH RELEVANCE: Project Narrative: Despite the emergence of guidelines and standard for management of acute post operative pain, unrelieved pain persists, driving increased costs of hospitalization. This work will perform a secondary data analysis of electronic health records, including nursing pain assessments and therapies (both pharmacologic and non-pharmacologic). We will also data mine for other implicit, previously unknown, useful information such as conditions that should trigger or escalate an alert for an acute pain management intervention.