1471-2288-8-72.fm

BMC Medical Research
Methodology
Research article
Use of the Oxford Handicap Scale at hospital discharge to predict
Glasgow Outcome Scale at 6 months in patients with traumatic
brain injury
Pablo Perel*, Phil Edwards, Haleema Shakur and Ian Roberts
Address: Department of Epidemiology and Population Health, London School of Hygiene & Tropical Medicine, London, UK Email: Pablo Perel* - pablo.perel@lshtm.ac.uk; Phil Edwards - phil.edwards@lshtm.ac.uk; Haleema Shakur - haleema.shakur@lshtm.ac.uk; Ian Roberts - ian.roberts@lshtm.ac.uk Received: 24 June 2008Accepted: 6 November 2008 BMC Medical Research Methodology 2008, 8:72
2008 Perel et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Background: Traumatic brain injury (TBI) is an important cause of acquired disability. In evaluating
the effectiveness of clinical interventions for TBI it is important to measure disability accurately.
The Glasgow Outcome Scale (GOS) is the most widely used outcome measure in randomised
controlled trials (RCTs) in TBI patients. However GOS measurement is generally collected at 6
months after discharge when loss to follow up could have occurred. The objectives of this study
were to evaluate the association and predictive validity between a simple disability scale at hospital
discharge, the Oxford Handicap Scale (OHS), and the GOS at 6 months among TBI patients.
Methods: The study was a secondary analysis of a randomised clinical trial among TBI patients
(MRC CRASH Trial). A Spearman correlation was estimated to evaluate the association between
the OHS and GOS. The validity of different dichotomies of the OHS for predicting GOS at 6
months was assessed by calculating sensitivity, specificity and the C statistic. Uni and multivariate
logistic regression models were fitted including OHS as explanatory variable. For each model we
analysed its discrimination and calibration.
Results: We found that the OHS is highly correlated with GOS at 6 months (spearman correlation
0.75) with evidence of a linear relationship between the two scales. The OHS dichotomy that
separates patients with severe dependency or death showed the greatest discrimination (C
statistic: 84.3). Among survivors at hospital discharge the OHS showed a very good discrimination
(C statistic 0.78) and excellent calibration when used to predict GOS outcome at 6 months.
Conclusion: We have shown that the OHS, a simple disability scale available at hospital discharge
can predict disability accurately, according to the GOS, at 6 months. OHS could be used to improve
the design and analysis of clinical trials in TBI patients and may also provide a valuable clinical tool
for physicians to improve communication with patients and relatives when assessing a patient's
prognosis at hospital discharge.
Trial Registration Number: ISRCTN74459797
(page number not for citation purposes) BMC Medical Research Methodology 2008, 8:72
http://www.biomedcentral.com/1471-2288/8/72 Background
form of the OHS was used in which moderate handicap Traumatic brain injury (TBI) is an important cause of and moderately severe handicap were combined.
acquired disability. In evaluating the effectiveness of clin- Although the OHS has been previously used in brain ical interventions for TBI it is important to measure disa- injury trials, its association with GOS at 6 months in TBI bility accurately. The Glasgow Outcome Scale (GOS) is patients has not been previously reported.[ the most widely used outcome measure in randomisedcontrolled trials (RCTs) in TBI patients.[ However, The aim of this paper is to describe the association because the GOS assesses how well patients function in between an early disability outcome (OHS), and a 6 their daily social interactions, it is only applicable after the months disability outcome (GOS). Specifically the objec- patient has been discharged from hospital.
Loss to follow up after hospital discharge is a common 1) Evaluate the correlation between OHS at hospital dis- problem in clinical trials in TBI and some amount of miss- ing data is often unavo] If an early outcomemeasure was available that could predict long term disa- 2) Evaluate different dichotomies of the OHS at hospital bility, it could be valuable for dealing with missing data, and might potentially be used as a surrogate outcome.
3) Evaluate the extent to which OHS at hospital discharge The MRC CRASH Trial was a large, randomised placebo controlled trial of the effects of a 48-hour infusion of cor-ticosteroids on death and disability, among 10,008 ad] Using data from this cohort of patients we have Potential predictor
previously identified hospital admission variables that The OHS (tables assessed at 14 days, hospital dis- accurately predict 6 months GO This cohort also charge or death (whichever occur first).
presents an opportunity to evaluate the predictive validityof an early disability outcome measure for TBI patients. A Variables that have previously been reported to be associ- modified version of the Oxford Handicap Scale (OHS) ated with the outcome were considered as potential con- was completed at hospital discharge and the GOS was founders and included in an adjusted model: age, completed at 6 months after injury. The OHS, which was Glasgow Coma Scale (GCS) at randomization, pupil reac- originally developed for stroke patients, comprises six cat- tivity, whether the patient sustained a major extra cranial egories: no symptoms, minor symptoms, minor handi- injury and computerised tomography (CT) scan cap, moderate handicap, moderately severe handicap, and Table 1: Original Oxford Handicap Scale and OHS used in the MRC CRASH Trial
Original OHS
Modified OHS used in CRASH
Categories
Categories
Minor symptoms that do not interfere with lifestyle Minor handicap, symptoms that lead to some restriction in lifestyle but do not interfere with Some restriction in lifestyle but independent the patient's capacity to look after himself Moderate handicap, symptoms that significantly restrict lifestyle and prevent totally independent Dependent but not requiring constant attention Moderately severe handicap, symptoms that clearly prevent independent existence though not needing constant attention Severe handicap, totally dependent patient requiring constant attention night and day Fully dependent requiring attention day and night (page number not for citation purposes) BMC Medical Research Methodology 2008, 8:72
http://www.biomedcentral.com/1471-2288/8/72 418 no dat a on GOS at 6 m ont hs Sa m p l e f o r o b j e ct i v e s 1 a n d 2 Flowchart of patients used in the analysis.
Outcome
sent waiver and others consent from a legal representative.
The outcome was GOS at 6 months. The GOS comprises We always adhered to these requirements.
five categories: death, persistent vegetative state, severedisability, moderate disability and good re Of 10,008 study participants enrolled in the MRC CRASH was dichotomised for analysis in the CRASH Trial into Trial, 99 (1%) had missing data on the OHS, 418 (4.2%) favourable outcome (good recovery or moderate disabil- had missing data on the GOS at 6 months, and 36 (0.3%) ity) and unfavourable outcome (severe disability, persist- had missing data for both OHS and GOS. A further 8 ent vegetative state or death). We created two further patients were excluded from analysis as they had a Glas- dichotomies: good recovery versus other outcomes, and gow Coma Scale (GCS) score of 15 at randomisation.
Analysis for objectives 1 and 2 were therefore performedusing data for 9,447 (94.4%) patients (fige The sample of patients
third objective (predictive validity of OHS among survi- The MRC CRASH trial was a large international double- vors), the 1,948 patients who died within 14 days of blind randomised placebo-controlled trial of the effect of admission were excluded and the analysis was based on early administration of a 48-h infusion of a corticosteroid data for the remaining 7,499 patients (figure (methylprednisolone) on the risk of death and disabilityafter TBI. The characteristics of the patients randomised, Analysis
and results of the trial have already been reported in de, Briefly, adults (aged 16 years or older) with a A cross-tabulation between the OHS and GOS categories head injury and a GCS of 14 or less were randomly allo- was performed. Their relation was assessed with the Spear- cated to commence either a 48 hour infusion of methyl- prednisolone or matching placebo within eight hours ofinjury; patients from 239 hospitals in 48 countries were randomised. All collaborating MRC CRASH investigators The validity of the different dichotomies of the OHS for were required to secure local ethics or research committee predicting GOS at 6 months was assessed by calculating approval before recruitment could begin. Patients with sensitivity, specificity and the c statistic (an equivalent clinically significant head injury are unable to give valid concept to area under the receiver operator characteristic informed consent. Local ethics committees set consent procedures for participating hospitals. Some allowed con- (page number not for citation purposes) BMC Medical Research Methodology 2008, 8:72
http://www.biomedcentral.com/1471-2288/8/72 (37.3%) patients were severely disabled or had died. Most A logistic regression model was first fitted including only deaths (84%) occurred within the first 14 days. OHS at 14 OHS as explanatory variables (model 1). A second model days and GOS at 6 months were highly correlated (Spear- was then fitted that also included demographic and clini- man rank correlation coefficient 0.75) and they showed a cal variables (model 2). Finally, a third model was fitted that included all variables in model 2, plus CT scan varia-bles. All the demographic, clinical and CT variables have OHS for predicting 6 months outcome
been previously reported as being independently associ- Five dichotomies of the OHS were consider ated with unfavourable outcome at 6 months.[ eachmodel we analysed its discrimination using the c statistic When their validity was assessed in relation to unfavoura- and calibration (graphically and with the Hosmer-Leme- ble outcome as defined by the GOS (severe disability or death), dichotomy D showed the highest discrimination(c statistic) with high specificity (Table We then estimated the positive predictive value (with 95%confidence intervals) of each OHS category for GOS at 6 Among survivors at hospital discharge the OHS showed a strong association with GOS at 6 months. The crude anal-ysis showed that patients who were fully dependent at 14 days had 24 higher odds of an unfavourable outcome at 6 General characteristics of the population
months. Although adjusting for known prognostic factors Taows the characteristics of the sample included in attenuated the strength of the association, OHS remained the analysis. At 14 days 1,863 (19%) were dependent and a strong predictor with a highly statistically significant test 1,948 patients had died (21%). At 6 months, 3,525 between Oxford Handicap Scale and unfavourable outcome (GOS) at 6 months Relationship between Oxford Handicap Scale and unfavourable outcome (GOS) at 6 months.
(page number not for citation purposes) BMC Medical Research Methodology 2008, 8:72
http://www.biomedcentral.com/1471-2288/8/72 Table 2: Glasgow Outcome Scale at 6 months by Oxford Handicap Scale at 14 days
Glasgow Outcome Scale at 6 months
Oxford Handicap Scale at 14 days
Good recovery
Severe disability
Total
Some restriction in lifestyle but independent Dependent but not requiring constant attention Fully dependent requiring attention day & night OHS showed very good discrimination and excellent cali- hospital discharge the OHS showed a very good discrimi- nation and excellent calibration when used to predictGOS outcome at 6 months.
Ta shows the prediction of different 6 months out-comes (as measured with GOS) according to the hospital Strengths and weakness of the study
discharge OHS. For example, a patient with minor symp- To our knowledge this is the first study that evaluated the toms at hospital discharge will have a probability of predictive validity of a simple scale for disability at hospi- approximately 67% of good recovery, 89% of good recov- tal discharge in TBI patients. The main strengths of our ery or moderate disability and 98% of survival at 6 study include the large sample size which ensures preci- sion in our estimates, and the inclusion of patients fromboth high and low & middle income countries, which Discussion
increases the generalizability of our conclusions. The Principal findings
main limitation is that the measurement of OHS was not We found that the OHS is highly correlated with GOS at 6 standardized between centres. However, because we months with evidence of a linear relationship between the would expect that any measurement error would result in two scales. The OHS dichotomy that separate patients non-differential misclassification, in general we would who were severely dependent or dead (dichotomy D) expect that the association reported would be underesti- showed the greatest discrimination. Among survivors at mated rather than overestimated. Finally, our study is the Table 3: Dichotomies of OHS for determining unfavourable outcome
Some restriction in lifestyle but independent Dependent but not requiring constant attention Fully dependent requiring attention day and night (page number not for citation purposes) BMC Medical Research Methodology 2008, 8:72
http://www.biomedcentral.com/1471-2288/8/72 Table 4: Validity of the Oxford Handicap Scale at 14 days for
Glasgow Outcome Scale at 6 months
first to report this association which should therefore beexamined in an external cohort of patients in order to con- Calibration of model 1.
Comparison with other studies
The incidence of unfavourable GOS outcome at 6 months
Conclusion
in our cohort was lower in comparison to one reported in We have shown that OHS is strongly related and predicts a series of TBI coho] However, unlike ours, most of accurately the GOS at 6 months. It may therefore be help- these cohorts had been restricted to severe TBI patients.
ful in tackling the problem of missing data in clinical trials The OHS has previously been used in RCTs of brain injury in TBI. It might also serve as a potential surrogate outcome patients, and Bamford et al. reported good inter-observer measure and this application should be explored in fur- agreement (a weighted kappa of 0.72] Ours is the first ther studies. If our findings are replicated, OHS could be study in TBI which has evaluated the relationship between a simple and useful outcome measure to use in trials in OHS and GOS. Nevertheless, previous studies have shown settings for which long term follow-up is problematic.
a good agreement between the Modified Rankin Scale (the Furthermore, OHS could be a useful variable to collect in scale from which the OHS was derived) and the rehabilitation trials in TBI patients to ensure that there isa similar distribution of disability among participants Table 5: Association between OHS and unfavourable outcome (GOS) among survivors
Model 1
Model 3
Some restriction in lifestyle but independent Dependent but not requiring constant attention Fully dependent requiring attention day & night Model 1: OHSModel 2: model 1 plus GCS, pupil reactivity, major extra-cranial injury and ageModel 3: model 2 plus CT findings (petechial haemorrhages, obliteration of the third ventricle or basal cisterns, subarachnoid bleed, midline shift, non evacuated haematoma) (page number not for citation purposes) BMC Medical Research Methodology 2008, 8:72
http://www.biomedcentral.com/1471-2288/8/72 Table 6: Prediction of three dichotomies of GOS at 6 months according to OHS
Good recovery or Moderate disability Minor symptoms
Some restriction in lifestyle but independent Dependent but not requiring constant attention Fully dependent requiring attention day & night Maas AI, Marmarou A, Murray GD, Teasdale SG, between groups at baseline. We have also shown that, among survivors, the OHS is able to predict disability at 6 J Neurotrauma 2007, 24(2):232-238.
Tilley BC, Marler J, Geller NL, Lu M, Legler J, Brott T, Lyden P, Grotta months and thus may provide a valuable clinical tool for physicians to improve communication with patients and relatives when assessing a patient's prognosis at hospital Stroke 1996,
27(11):2136-2142.
Pre-publication history
Competing interests
The pre-publication history for this paper can be accessed The authors declare that they have no competing interests.
Authors' contributions
ww.biomedcentral.com/1471-2288/8/72/prepub PP designed the study, performed the analysis and pre-pared the manuscript. IR conceived the study, revised anddrafted the manuscript. PE and HS revised and drafted themanuscript.
Acknowledgements
The authors express their gratitude to all of the study participants and all
the collaborators from the MRC CRASH Trial. We also want to thank
Taemi Kawahara for helping with the tables and figures.
References
1.
Bullock MR, Merchant RE, Choi SC, Gilman CB, Kreutzer JS, Marma-
rou A, Teasdale GM: Neurosurg Focus 2002, 13(1):ECP1.
Dickinson K, Bunn F, Wentz R, Edwards P, Roberts I: BMJ 2000, 320(7245):1308-1311.
scientist can read your work free of charge "BioMed Central will be the most significant development for 2005, 365(9475):1957-1959.
disseminating the results of biomedical researc h in our lifetime." CRASH Trial Collaborators: Predicting outcome after trau-
matic brain injury: practical prognostic models based on
large cohort of international patients. BMJ 2008, 12:12.
Bamford JM, Sandercock PA, Warlow CP, Slattery J: available free of charge to the entire biomedical community Stroke 1989, 20(6):828.
peer reviewed and published immediately upon acceptance Jennett B, Bond M: Lancet 1975, 1(7905):480-484.
cited in PubMed and archived on PubMed Central Lancet 2004, 364(9442):1321-1328.
(page number not for citation purposes)

Source: http://www.eurotherm3235trial.eu/Data/upload/Files/Perel%20et%20al%20Use%20of%20Oxford%20handicap%20scale%20at%20hosp%20discharge%20to%20predict%20GOS%20at%206%20months%20in%20TBI%202008(1).pdf