Predicting Hospital Re-Admissions from Nursing Care Data of Hospitalized Patients

Abstract

Readmission rates in the hospitals are increasingly being used as a benchmark to determine the quality of healthcare delivery to hospitalized patients. Around three-fourths of all hospital re-admissions can be avoided, saving billions of dollars. Many hospitals have now deployed electronic health record (EHR) systems that can be used to study issues that trigger readmission.However, most of the EHRs are high dimensional and sparsely populated, and analyzing such data sets is a Big Data challenge. The effect of some of the well-known dimension reduction techniques is minimized due to presence of non-linear variables. We use association mining as a dimension reduction method and the results are used to develop models, using data from an existing nursing EHR system, for predicting risk of re-admission to the hospitals. These models can help in determining effective treatments for patients to minimize the possibility of re-admission, bringing down the cost and increasing the quality of care provided to the patients. Results from the models show significantly accurate predictions of patient re-admission.

Secondary use of standardized nursing care data for advancing nursing science and practice: a systematic review

“…Monsen et al 64 showed how specific interventions delivered by public health nurses during home visits were associated with decreased risks for pregnant women and their families suffering from social disadvantages and poverty. Lodhi et al 38 showed that SNT-coded nursing care plans are valuable in predicting hospital readmissions, which may help practitioners develop strategies to identify at-risk patients and potentially reduce healthcare costs in the future.…”

“…Twenty-six (58%) studies analyzed nursing data coded with NANDA-I, [26][27][28] NOC, 29,30 NIC, [31][32][33] a combination of these terminologies (eg, NIC and NOC), [34][35][36] or the NNN set. [37][38][39][40][41][42][43][44][45][46][47][48][49][50][51] Fourteen (31%) studies used Omaha System, [52][53][54][55][56][57][58][59][60][61][62][63][64][65] while 4 (9%) used ICNP. [66][67][68] One (2%) study compared nursing data coded with NANDA-I and ICNP from different EHRs.…”

“…Yao et al 44 identified nursing diagnoses and interventions associated with changes in outcomes that could be used as an indicator of a shift from standard nursing care to palliative care. Lodhi et al 38 created predictive models for hospital readmission using patient's age, length of stay, nurse years of experience, time of admission and discharge, and pain outcomes ratings. Last, Stifter et al 47 used nursing diagnoses, interventions, and outcomes to determine presence of pressure ulcers in hospitalized patients, and then examined their association with nurse staffing (eg, nurse years of experience, education, shift length, work pattern) and nurse continuity.…”

“…26,[31][32][33]45,46,57,67 The majority (16 of 30, 53%) of studies analyzed care plans and other nursing data. 28,30,[36][37][38][39][40][41]43,44,47,54,59,[63][64][65] Table 2 shows 22 SNT-coded nursing datasets. Ten (45%) of 22 were retrieved from at least 2 different institutions/clinics.…”

Analysis of Feature Extraction Methods for Prediction of 30-Day Hospital Readmissions

“…11,12 Other models are specifically designed for specific types of care. 13,14 There have been many different techniques used by models trying to predict hospital readmissions, such as support vector machines (SVMs), 15,16 classification trees (CTs), 7,15 and random forests (RFs), 9,17 along with techniques such as the LACE index, 18 which scores a patient's risk for readmission based on certain factors. One such method that has been used in predicting hospital readmission is Cox hazard regression analysis.…”