Investigating the feasibility of crisis-discharge decision-support to reduce readmission rates at a psychiatric ward.

Date
2016-12
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: The pressure on the availability of beds in South African psychiatric hospitals is high. In response, the Western Cape has implemented a \crisis" discharge policy. This policy is different from the planned short stay practice as the patients are discharged earlier than what is clinically ideal to allow people on the waiting list to be admitted. The crisis-discharge policy therefore strives to optimise the combined healthcare outcome for the patient population (including those currently in the ward and those on the waiting list for admission). Discharge and admission decisions are informed by reviewing clinical indicators of the patient population. A previous study conducted at Stikland Psychiatric Hospital, which is also the institution where this research was undertaken, reported crisis-discharge to be a significant predictor of an increased risk for readmission. This suggests that implementing a crisis-discharge policy to alleviate the pressure on available beds, may in fact exacerbate the scenario. Currently, unaided decision making is implemented by the clinical psychiatrists to solve this combinatorial optimisation problem and it is therefore unlikely that the daily decisions are optimal. This study investigates readmission at Stikland Psychiatric Hospital, specifically at the acute male inpatient ward to (i) determine whether variables exist that indicate that certain patients within this population have a higher risk of requiring readmission after a crisis-discharge; and, if such variables do exist, (ii) to determine the predictive capability of these variables with a view of recommending the feasibility of a decision-support system for crisis-discharge at the male inpatient ward. Various patient variables such as age, diagnosis, place of follow-up and substance use are analysed. Basic descriptive methods, biostatistics and data mining were employed to analyse the data. Predictive models were developed and evaluated using, amongst others, classi cation and regression trees and random forests. The research was conducted with continuous input from clinical subject matter experts. The most important statistically significant variables pertaining to the risk of readmission are the diagnosis, whether a patient belongs to a community after-care programme, and the area that a patient originates from. Direct admissions and schizophrenic patients were found to be twice as likely to be readmitted as patients who are not from these groups. The schizo-affective and bipolar diagnostic groups are about three times as likely to be readmitted compared to patients who are not from these diagnostic groups. The substance induced psychosis diagnostic variable, and a community programme variable indicated that patients were less than half as likely to require readmission. These results are some of the insights that are presented in this research project. The best-performing predictive model is able to classify whether patients would require readmission following a crisis-discharge with average accuracy of 70%. Based on these findings, further research towards to the development of a crisis-discharge decision-support tool is recommended.
AFRIKAANSE OPSOMMING: Die aanvraag na beskikbare beddens in Suid-Afrikaanse psigiatriese hospitale is hoog. Dit het gelei tot die implementering van 'n krisis-ontslag beleid in die Wes-Kaap. Hierdie beleid verskil van die beplande korter-hospitaalvertoeftyd aangesien 'n pasiënt vroeër ontslaan word as wat klinies ideaal is sodat 'n persoon op die waglys toegelaat kan word. Die krisis-ontslag beleid poog dus om die gekombineerde gesondheidsorg uitkoms van die totale pasiëntpopulasie te optimeer (dit sluit die pasiënte in die saal asook die op die waglys in). Besluite rakende die ontslaan en toelating van pasiënte word geneem deur kliniese veranderlikes van die pasiëntpopulasie in ag te neem. 'n Vorige studie gedoen by die Stikland Psigiatriese Hospitaal, dieselfde instansie waarby hierdie navorsing gedoen is, het bevind dat krisis-ontslag 'n beduidende bepaler van hertoelating is. Dit dui daarop dat 'n krisis-ontslag beleid nie die druk op die aantal beskikbare beddens verlig nie, maar dit egter kan vererger. Tans word die kombineerde optimaliseringsprobleem opgelos deur die kliniese psigiaters wat op 'n daaglikse basis besluite moet neem, en gevolglik is dit onwaarskynlik dat die besluite wel optimaal is. Hierdie navorsingsprojek ondersoek hertoelating, spesifiek by die akute manssaal van Stikland Psigiatriese Hospitaal, om vas te stel of (i) daar veranderlikes bestaan wat daarop dui dat sekere pasiënte van hierdie populasie 'n hoër risiko het om hertoegelaat te word na 'n krisis-ontslag, en, indien diesulke veranderlikes voorkom, (ii) wat die voorspellingsvermoë van hierdie veranderlikes is om sodoende die moontlikheid van 'n besluitsteunstelsel vir krisis-ontslag vas te stel. Pasiënt veranderlikes wat ondersoek word is onder andere ouderdom, diagnose, opvolg en dwelmgebruik. Die data is geanaliseer met beskrywende statistiek, verskeie biostatistiese en data mining metodes. Klassifikasie en regressie bome en random forests is onder andere gebruik om voorspellingsmodelle te ontwikkel en te evalueer. Die navorsing het deurlopende kontaksessies met kliniese kundiges behels. Die belangrikste statisties beduidende veranderlikes, wat die risiko vir hertoelating benadruk, is die gebied vanwaar 'n pasiënt kom, diagnose en of 'n pasiënt aan 'n gemeenskapsopvolg-program behoort. Direkte toelatings{ en skisofreniese pasiënte het 'n twee maal groter waarskynlikheid vir hertoelating as pasiënte wat nie onder hierdie twee groepe klassifiseer nie. Die skiso-affektiewe{ en bipolêre diagnose groep is ongeveer drie keer meer waarskynlik vir hertoelating as pasiënte wat nie in hierdie diagnose-groepe val nie. Daar is ook gevind dat die substansgeïnduseerde psigose veranderlike en pasiënte wat behoort aan 'n spesifieke gemeenskapsopvolg-program meer as die helfte minder kans het om hertoegelaat te word. Hierdie resultate verteenwoordig van die bevindinge wat bespreek word in hierdie navorsingsprojek. Die beste voorspellingsmodel het met 'n gemiddelde akkuraatheid van 70% voorspel of krisisontslag pasiënte weer hertoegelaat gaan word. Hierdie bevindinge lei dan tot die aanbeveling dat verdere navorsing gedoen moet word vir die ontwikkeling van 'n krisis-ontslag besluitsteunprogram.
Description
Thesis (MEng)--Stellenbosch University, 2016.
Keywords
Psychiatric hospitals -- Admission and discharge, UCTD, Psychiatric hospital patients, Health services administration
Citation