Can the simple clinical score usefully predict the mortality risk and length of stay for a recently admitted patient?
Minh T. Nguyen A , Richard J. Woodman B , Paul Hakendorf B C , Campbell H. Thompson A E and Jeff Faunt DA Discipline of Medicine, University of Adelaide, North Terrace, Adelaide, SA 5005, Australia. Email: thiennguyen223@hotmail.com
B Flinders Centre for Epidemiology and Biostatistics, School of Medicine, Flinders University, Sturt Road, Bedford Park, SA 5042, Australia. Email: richard.woodman@flinders.edu.au
C Clinical Epidemiology, Flinders Medical Centre, Flinders Drive, Bedford Park, SA 5042, Australia. Email: Paul.Hakendorf@health.sa.gov.au
D Department of General Medicine, Royal Adelaide Hospital, North Terrace, Adelaide, SA 5000, Australia. Email: Jeff.Faunt@health.sa.gov.au
E Corresponding author. Email: campbell.thompson@adelaide.edu.au
Australian Health Review 39(5) 522-527 https://doi.org/10.1071/AH14123
Submitted: 29 July 2014 Accepted: 4 February 2015 Published: 30 March 2015
Abstract
Objectives The aim of the present study was to determine whether an aggregate simple clinical score (SCS) has a role in predicting the imminent mortality and in-hospital length of stay (LOS) of newly admitted, acutely unwell General Medical in-patients.
Methods Data were collected prospectively from adult patients admitted through an Acute Medical Unit between February and August 2013. Using logistic regression analysis before and after adjustment for age, the SCS was assessed for its association with LOS and mortality, including 30-day mortality, just for those patients for full resuscitation. Changes in sensitivity and specificity after adding SCS to age as a predictor, as well as the change in the net reclassification index, were determined using the predicted probabilities from the logistic regression models.
Results The SCS was superior to age in predicting mortality of any patient within 30 days. It did not assist in predicting 30-day mortality for those patients who were for full resuscitation. The ability of the SCS to predict long stay (>72 h) remained relatively low (64%) and was inferior to published rates achieved by bedside clinician assessment (74%–82%).
Conclusion There was no useful prospective role for the SCS in predicting LOS and mortality of in-patients newly admitted to a General Medicine service.
What is known about the topic? After their presentation to the emergency department, care efficiency is improved by the ‘streaming’ of patients according to their risk of imminent deterioration and their likelihood of being a long-stay patient. Although streaming is currently effected by bedside assessment of the patient, an accepted aggregate assessment score may assist disposition decisions.
What does this paper add? Bedside assessment of each patient still offers the most accurate method for identifying the long-stay patient. The SCS, good at predicting 30-day mortality of all new admissions, is not useful for predicting the death of those admissions who are for full resuscitation.
What are the implications for practitioners? When deciding admitted patients’ disposition on leaving the emergency department, a simple aggregate score based on patient physiology, comorbidity and functionality has little to offer practitioners beyond knowledge of each patient’s age.
References
[1] Kellett J, Deane B. The simple clinical score predicts mortality for 30 days after admission to an acute medical unit. Q J Med 2006; 99 771–81.| The simple clinical score predicts mortality for 30 days after admission to an acute medical unit.Crossref | GoogleScholarGoogle Scholar | 1:STN:280:DC%2BD28njs1aisA%3D%3D&md5=db336beed7f27feb9120f21fa585d8fdCAS |
[2] Li JYZ, Yong TY, Hakendorf P, Roberts S, O’Brien LT, Sharma Y, Ben-Tovim D, Thompson CH. The simple clinical score is associated with mortality, length of stay and readmission rate of acute general medical admissions to an Australian hospital. Int Med J 2012; 42 160–5.
| The simple clinical score is associated with mortality, length of stay and readmission rate of acute general medical admissions to an Australian hospital.Crossref | GoogleScholarGoogle Scholar | 1:CAS:528:DC%2BC38XhtlWqu7jO&md5=da576d0c5ed210c688e5df84e4848081CAS |
[3] Kellett J, Deane B. What diagnoses may make patients more seriously ill than they first appear? Mortality according to the simple clinical score risk class at the time of admission compared with the observed mortality of different ICD9 codes identified on death or discharge. Eur J Intern Med 2009; 20 89–93.
| What diagnoses may make patients more seriously ill than they first appear? Mortality according to the simple clinical score risk class at the time of admission compared with the observed mortality of different ICD9 codes identified on death or discharge.Crossref | GoogleScholarGoogle Scholar | 19237100PubMed |
[4] McNeill D, Mohapatra B, Li J, Spriggs D, Ahamed S, Gaddi Y, Hakendorf P, Ben-Tovim DI, Thompson CH. Quality of resuscitation orders in general medical patients. Q J Med 2012; 105 63–8.
| Quality of resuscitation orders in general medical patients.Crossref | GoogleScholarGoogle Scholar | 1:STN:280:DC%2BC38%2FlvFahsg%3D%3D&md5=8ac7e0cd30d96b8910bbe8c651f040b4CAS |
[5] O’Brien D, Williams A, Blondell K, Jelinek GA. Impact of streaming ‘fast track’ emergency department patients. Aust Health Rev 2006; 30 525–32.
| Impact of streaming ‘fast track’ emergency department patients.Crossref | GoogleScholarGoogle Scholar | 17073548PubMed |
[6] Dent AW, Weiland TJ, Vallender L, Oettel NE. Can medical admission and length of stay be accurately predicted by emergency staff, patients or relatives? Aust Health Rev 2007; 31 633–41.
| Can medical admission and length of stay be accurately predicted by emergency staff, patients or relatives?Crossref | GoogleScholarGoogle Scholar | 17973623PubMed |
[7] Dinh MM, Bein KJ, Byrne CM, Gabbe B, Ivers R. Deriving a prediction rule for short-stay admission in trauma patients admitted at a major trauma centre in Australia. Emerg Med J 2014; 31 263–7.
| Deriving a prediction rule for short-stay admission in trauma patients admitted at a major trauma centre in Australia.Crossref | GoogleScholarGoogle Scholar | 23407379PubMed |
[8] Pencina MJ, D’Agostino RB, D’Agostino RB, Vasan RS. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med 2008; 27 157–72.
| Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond.Crossref | GoogleScholarGoogle Scholar | 17569110PubMed |
[9] Shanmuganathan N, Li JYZ, Yong TY, Hakendorf P, Ben-Tovim DI, Thompson CH. An audit of resuscitation orders and their relevance to patients’ clinical outcomes. Q J Med 2011; 104 485–8.
| An audit of resuscitation orders and their relevance to patients’ clinical outcomes.Crossref | GoogleScholarGoogle Scholar | 1:STN:280:DC%2BC3MrjslKgsQ%3D%3D&md5=50489ad29340c46c756b99b08f20f6e6CAS |
[10] Subbe CP, Gauntlett W, Kellett JG. Collaborative audit of risk evaluation in medical emergency treatment (CARE-MET 1): an international pilot. Eur J Intern Med 2010; 21 222–5.
| Collaborative audit of risk evaluation in medical emergency treatment (CARE-MET 1): an international pilot.Crossref | GoogleScholarGoogle Scholar | 1:STN:280:DC%2BC3czmt1artg%3D%3D&md5=3187024c46c8b44de7afca021a8453bfCAS | 20493426PubMed |
[11] Paterson R, MacLeod DC, Thetford D, Beattie A, Graham C, Lam S. Prediction of in-hospital mortality and length of stay using an early warning score system: clinical audit. Clin Med 2006; 6 281–4.
| Prediction of in-hospital mortality and length of stay using an early warning score system: clinical audit.Crossref | GoogleScholarGoogle Scholar | 1:STN:280:DC%2BD28vgs1entA%3D%3D&md5=f461c8cb3e238c292914a9b5c91a5e86CAS | 16826863PubMed |
[12] McGaughey J, Alderdice F, Fowler R, Kapila A, Mayhew A, Moutray M. Outreach and early warning systems (EWS) for the prevention of intensive care admission and death of critically ill adult patients on general hospital wards. Cochrane Database Syst Rev 2007; 3 CD005529
| Outreach and early warning systems (EWS) for the prevention of intensive care admission and death of critically ill adult patients on general hospital wards.Crossref | GoogleScholarGoogle Scholar | 17636805PubMed |
[13] Damiani G, Pinnarelli L, Sommella L, Vena V, Magrini P, Ricciardi W. The short stay unit as a new option for hospitals: a review of the scientific literature. Med Sci Monit 2011; 17 SR15–19.
| The short stay unit as a new option for hospitals: a review of the scientific literature.Crossref | GoogleScholarGoogle Scholar | 21629205PubMed |
[14] Lucas BP, Kumapley R, Mba B, Nisar I, Lee K, Otori-Ntow S, Borkowsky S, Asmar A, Lewis T, Bienias JL. A hospitalist-run short stay unit: feature that predict length-of-stay and eventual admission to traditional inpatient services. J Hosp Med 2009; 4 276–84.
| A hospitalist-run short stay unit: feature that predict length-of-stay and eventual admission to traditional inpatient services.Crossref | GoogleScholarGoogle Scholar | 19504489PubMed |
[15] Yong TY, Li JYZ, Roberts S, Hakendorf P, Ben-Tovim DI, Thompson CH. The selection of acute medical admissions for a short stay unit. Intern Emerg Med 2011; 6 321–7.
| The selection of acute medical admissions for a short stay unit.Crossref | GoogleScholarGoogle Scholar | 21161437PubMed |
[16] Russell PT, Hakendorf P, Thompson CH. A general medical short stay unit is not more efficient than a traditional model of care. Med J Aust 2014; 200 482–4.
| A general medical short stay unit is not more efficient than a traditional model of care.Crossref | GoogleScholarGoogle Scholar | 24794612PubMed |
[17] Downing H, Scott C, Kelly CA. Evaluation of a dedicated short-stay unit for medical admission. Clin Med 2008; 8 18–20.
| Evaluation of a dedicated short-stay unit for medical admission.Crossref | GoogleScholarGoogle Scholar | 18335661PubMed |
