Abstract
Objective
To explore the added value of posttraumatic confusional state (PTCS) duration for outcome prediction, in patients with moderate and severe traumatic brain injury (TBI).
Design
Neurosurgical inception cohort study with follow-up 12 months postinjury.
Setting
Regional trauma center.
Participants
Patients aged ≥16 years admitted with moderate or severe TBI (Glasgow Coma Scale score 9-13 and 3-8), who survived the acute phase. Three hundred ninety-five patients completed follow-up and had a valid PTCS duration estimation; 75% were men.
Interventions
Not applicable.
Main Outcome Measures
PTCS duration was pragmatically categorized into weekly intervals, primarily through retrospective review of medical records. In addition to PTCS duration, predictors included age, sex, the admission Glasgow Coma Scale score, pupillary dilatation, and the worst Rotterdam computed tomography score. The outcome was assessed using the Glasgow Outcome Scale–Extended (GOSE). Uni- and multivariable binary logistic regression analyses were performed to explore predictive models with and without PTCS duration. The dependent variable GOSE was dichotomized using several cutoffs: GOSE scores ≤7, ≤6, ≤5, and ≤4.
Results
The GOSE score (with lower scores indicating worse function) decreased with longer PTCS duration in patients with moderate and severe TBI. PTCS duration was a significant predictor of most outcomes and the only significant predictor in the multivariable models for severe TBI. Adding PTCS duration as a covariate improved the fit of the multivariable models, particularly in severe TBI. When PTCS lasted <28 days, a GOSE score of 1-4 was observed in only 3% of cases.
Conclusions
The strong association between PTCS duration and outcomes demonstrates the benefit of estimating PTCS in hospital and rehabilitation settings. Moreover, PTCS holds promise as a modifier of the TBI severity classification.
KEYWORDS: Adult, Craniocerebral trauma, Glasgow Coma Scale, Glasgow Outcome Scale, Posttraumatic amnesia, Posttraumatic confusional state, Rehabilitation
Accurate prediction of functional outcome after traumatic brain injury (TBI) can support early clinical decision-making and provide predictability for relatives. To date, the most used predictors include age, clinical indicators of severity (typically the Glasgow Coma Scale [GCS] score and pupillary response), secondary insults, and head computed tomography (CT) findings.1,2 However, in patient populations with moderate and severe TBI, these predictors largely fail to explain the variation in outcome,3 which calls for better prognostic models and refined classifications of injury severity.
After TBI, patients typically go through phases of recovery. One phase is characterized by confusion, emerging from more severe disorders of consciousness (ie, coma, vegetative state, and minimally conscious state).4 Several terms for this phase are used interchangeably; the most common are posttraumatic amnesia and posttraumatic confusional state (PTCS).5 Posttraumatic amnesia was the first term in use,6,7 and later PTCS was proposed,8 to more adequately cover neurobehavioral characteristics, for example, disturbed attention, disorientation, behavioral fluctuation, and disturbed sleep-wake cycle.4 These characteristics, in addition to amnesia, are associated with PTCS. Throughout this paper, PTCS will be used.
The duration of PTCS has been found to be associated with functional outcome.9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 However, most previous studies of PTCS have included only highly selected patient populations, often from rehabilitation centers,9,10,12,14,16,17,20, 21, 22, 23, 24, 25 likely because standardized assessment of PTCS is difficult to implement in the acute hospital setting. It is nonetheless likely that clinical descriptions of the patient's level of consciousness are available in the medical records. Such descriptions could be used for a pragmatic estimation of PTCS duration and thereby improve outcome prediction. Further, substantial differences in both PTCS duration and outcome have been found between moderate and severe TBI.26 Yet, the literature on the prognostic value of PTCS duration is limited to studies analyzing moderate and severe TBI together12,18,21,22 and only including severe TBI.9, 10, 11,19,20,27 These gaps warrant further exploration of the predictive ability of PTCS duration in hospital settings, analyzed separately for moderate and severe TBI.
In this 15-year inception cohort study of survivors of moderate and severe TBI, the objectives were to explore the added value of pragmatically estimated PTCS duration in prognostic models, using different cutoffs for predicting worse outcomes.
Methods
Setting and study period
This study was conducted from October 1, 2004 to October 1, 2019, at St. Olavs Hospital, Trondheim University Hospital in Central Norway. The region, which includes 2 counties, has a total population of 729,452 inhabitants (2019). St. Olavs Hospital, a level I/II trauma center, serves as the regional referral facility for all neurosurgical procedures and a general hospital for 317,462 of inhabitants in the region (2019).
Participants
The current study comprises patients aged ≥16 years, surviving the acute phase. Patients are included in the Trondheim moderate and severe TBI study, an ongoing prospective cohort study comprising all patients with a moderate and severe TBI admitted to the St. Olavs Hospital within 72 hours.28 Moderate TBI is defined by an admission GCS score of 9-13 and severe TBI by a GCS score of ≤8. All patients are invited for follow-up, except those who live in foreign countries, do not speak a Scandinavian language or English, or have severe psychosocial strains (fig 1).
Fig 1.
Flowchart of patients aged ≥16 years with moderate and severe TBI admitted to hospital between October 1, 2004 and October 1, 2019.
Study procedures in the Trondheim moderate and severe TBI study
Residents in neurosurgery and a study nurse identified and prospectively included eligible patients. The hospital's trauma registry was also regularly checked. Since 2015, a study nurse has screened referrals to head CT scans weekly. Acute phase and discharge data were recorded prospectively. Follow-up at 12 months postinjury was conducted either through visits (20%) or by telephone (80%) by trained interviewers. In cases where the self-reported outcome could be influenced by the patient's impaired cognition or awareness, a caregiver or next of kin was also interviewed (51%).
Study variables
Demographic variables included age at injury, biological sex (male or female), and work/study status. The work/study status was categorized as yes if a patient had a paid position or was registered as a student. Retirement was defined as receiving a retirement pension from the national insurance.
Comorbidities diagnosed before TBI were recorded as present if they affected daily functioning. These were further categorized as none, substance abuse, psychiatric disorder, neurologic disease, developmental disorder, heart/lung disease, cancer, several diseases/disorders (more than 1), and others.
The external cause of injury was categorized as falls, traffic collisions, violence, gunshot injury, and others (including sports-related injuries). Pupillary dilatation was noted as yes if dilatation was present in 1 or both pupils. The GCS score was recorded at admission, with the last nonsedated GCS score used in cases of prehospital intubation. If severe intoxication or sedation occurred before obtaining the first GCS score, the admission GCS score was considered unreliable. Head CT scans were reviewed by residents or consultants in radiology or neurosurgery, who were blinded to the outcome. The CT examination with the worst findings was used. The Rotterdam CT classification, a scoring system from 1-6, evaluates the presence of basal cisterns, midline shifts, and various mass lesions.29 The Rotterdam CT score was used because it is ordinal and has better discriminative ability than the Marshall CT classification, because of the inclusion of individual CT parameters.29
PTCS duration was assessed using several approaches: (1) the most common method was a review of day-to-day notes in the medical records, typically written by nurses, physiotherapists, and physicians. Clinical descriptions of behavior indicating orientation, memory, and agitation were evaluated. The resolution of PTCS was considered likely if the patient was described as oriented, having regained memory, and not being agitated for at least 2 consecutive days. (2) In some patients, either the 100-item Galveston Orientation and Amnesia Test30 or the Orientation Log31 was used for prospective assessment. In these cases, the first of 3 consecutive days with a Galveston Orientation and Amnesia Test score ≥75 or an Orientation Log score ≥25 indicated resolution of PTCS. These were typically patients admitted to a rehabilitation ward while still in PTCS. (3) In a minority of cases, patients and their relatives were asked at the 6-month follow-up when the patient became oriented, was no longer agitated or confused, and had regained day-to-day memory. Typical questions asked to family members included “Can you remember when he/she was able to remember events from 1 day to another?” and “Can you remember when he/she was no longer confused?” If patients and their families could not provide a definite answer about the PTCS duration and no other source of information (eg, medical records) could either, the duration was set to “unknown.” In all cases, PTCS duration was categorized into weekly intervals: 0-7 days, 8-14 days, 15-21 days, 22-28 days, and >28 days. Categorizing PTCS into weekly intervals has previously been shown to efficiently determine injury severity.32 Our methods for PTCS estimation (ie, interviews and medical records) did not allow for a more fine-graded assessment, such as the exact number of days. If the PTCS duration could not be assigned to a specific week interval, it was considered unknown.
TBI-related functional outcome at 12 months postinjury was assessed using the Glasgow Outcome Scale–Extended (GOSE).33 The scale ranges from 1 (death) to 8 (full recovery with only minor symptoms not affecting daily functioning) (Supplemental table S1, available online only at http://www.archives-pmr.org/). Previous studies have used different cutoffs for worse outcomes,9,34,35 depending on the research question and study population. In this study, 4 different cutoffs for worse outcome were analyzed for both moderate and severe TBI: GOSE scores ≤7, ≤6, ≤5, and ≤4.
Statistical analyses
Data were summarized using means and SD for continuous variables (age and the GCS score) and frequencies and percentages for categorical variables. The functional outcome measure (GOSE) was reported with the median and interquartile range. The relationship between PTCS duration and the GOSE score was assessed using the Spearman correlation. The association between demographic-/injury-related variables and functional outcome was explored with univariable and multivariable binary logistic regression analyses, with the outcome as the dependent variable.
The independent variables included age and GCS score (as continuous variables), and PTCS duration (0-7d, 8-14d, 15-28d, and >28d), sex, pupillary dilatation, and the Rotterdam CT score (scores 1-2, 3, and 4-6) as categorical variables. Because of the small number of cases, the PTCS duration categories 15-21 days and 22-28 days were combined in the analyses. Similarly, the Rotterdam CT scores were combined because of a few cases in some categories.
Likelihood ratio tests and model performance measures (Nagelkerke R2, Akaike information criterion [AIC], and Bayesian information criterion [BIC]) were reported to compare multivariable models without and with PTCS duration. Binary logistic regression analyses were not suitable for the GOSE score ≤4 as the dependent variable because of the distribution of PTCS durations across the cutoff. Instead, the sensitivity, specificity, and positive and negative predictive values of PTCS duration >28 days were presented. A significance level of 0.05 was used.
Missing data were handled using available case analyses; patients were excluded from analysis if at least one of the variables included in that analysis was missing (ie, comorbidity, work/study status, cause of injury, and the GCS score at admission). Multiple imputations of missing data were not considered necessary because only a few patients had missing values for independent variables in the multivariable binary logistic regression analyses (ie, the GCS score, n=17 [4%] and pupillary dilatation, n=10 [3%]).
All analyses were performed using the IBM SPSS version 29 software.
Ethics
The Regional Committee for Medical Research Ethics has approved this study (2009/2328). Surviving patients gave their informed consent, or their relatives did so if the patient was incapacitated or underage. A waiver of consent was approved for deceased patients and those not planned for follow-up.
Results
Patient characteristics
Out of the 533 surviving patients, 2 declined registration, 63 were not planned for follow-up, 24 had a missing or no reliable GOSE score, and 49 had an unknown PTCS duration. Ultimately, 395 patients were included in the study (see fig 1). Of these, 56% had moderate TBI and 44% had severe TBI. The mean age was 44 years; 48 years for moderate and 39 years for severe TBI, and 75% were men. At the time of injury, 15% had preexisting health conditions (table 1). Most patients were discharged either to another hospital (44%) or a hospital-based rehabilitation department (28%). The discharge destinations and respective lengths of stay in the hospital are shown in Supplemental Table S2 (available online only at http://www.archives-pmr.org/).
Table 1.
Injury-related data and demographics of 395 included patients.
| Variable | All Included Patients n=395 |
Moderate TBI n=222 |
Severe TBI n=173 |
|---|---|---|---|
| Age at injury | |||
| Mean (SD) | 44 (20) | 48 (20) | 39 (18) |
| Sex, n (%) | |||
| Male | 295 (75) | 157 (71) | 138 (80) |
| Female | 100 (25) | 65 (29) | 35 (20) |
| Comorbidity, n (%)* | |||
| None | 321 (85) | 177 (84) | 144 (86) |
| Substance abuse | 26 (7) | 15 (7) | 11 (6) |
| Psychiatric disorder | 10 (3) | 3 (1) | 7 (4) |
| Neurologic disease | 13 (3) | 10 (5) | 3 (2) |
| Developmental disorder | ≤3 | ≤3 | 0 |
| Heart/lung disease | 0 | 0 | 0 |
| Cancer | 0 | 0 | 0 |
| Several diseases/disorders | ≤ 3 | ≤ 3 | ≤ 3 |
| Other | ≤ 3 | ≤ 3 | 0 |
| Work/study status before injury, n (%)† | |||
| Yes | 274 (72) | 145 (67) | 129 (79) |
| No | 54 (14) | 33 (15) | 21 (12) |
| Retired | 53 (14) | 39 (18) | 14 (9) |
| External cause of injury, n (%)‡ | |||
| Fall | 172 (44) | 108 (50) | 64 (37) |
| Traffic collision | 181 (47) | 88 (41) | 93 (54) |
| Inside car | 88 (49) | 29 (33) | 59 (64) |
| On bike | 39 (22) | 32 (36) | 7 (8) |
| Pedestrian | 21 (12) | 9 (10) | 12 (13) |
| On motorbike | 24 (13) | 13 (15) | 11 (12) |
| Other | 8 (4) | 5 (6) | 3 (3) |
| Assault | 13 (3) | 6 (3) | 7 (4) |
| Other, including sport accidents | 21 (5) | 14 (6) | 7 (4) |
| GCS score at admission | |||
| Mean (SD)§ | 9 (4) | 12 (1) | 5 (2) |
| Pupillary dilatation present, n (%) | 53 (13) | 11 (5) | 42 (24) |
| Worst Rotterdam CT score, n (%) | |||
| 1 | 12 (3) | 7 (3) | 5 (3) |
| 2 | 106 (27) | 66 (30) | 40 (23) |
| 3 | 173 (44) | 111 (50) | 62 (36) |
| 4 | 65 (17) | 30 (14) | 35 (20) |
| 5 | 31 (8) | 6 (3) | 25 (15) |
| 6 | 8 (2) | 2 (1) | 6 (4) |
| 12 mo GOSE score, n (%)|| | |||
| Median (25th and 75th percentile) | 6 (5, 8) | 7 (6, 8) | 5 (4, 7) |
| 8 | 125 (32) | 94 (44) | 31 (18) |
| 7 | 62 (16) | 42 (19) | 20 (12) |
| 6 | 66 (17) | 34 (16) | 32 (19) |
| 5 | 61 (16) | 22 (10) | 39 (23) |
| 4 | 18 (5) | 6 (3) | 12 (7) |
| 3 | 47 (12) | 17 (8) | 30 (17) |
| 2 | 2 (1) | 0 | 2 (1) |
| 1 | 8 (2) | 1 (1) | 7 (4) |
| PTCS duration, n (%) | |||
| 0-7 d | 161 (41) | 130 (59) | 31 (18) |
| 8-14 d | 47 (12) | 28 (13) | 19 (11) |
| 15-21 d | 25 (6) | 14 (6) | 11 (6) |
| 22-28 d | 35 (9) | 15 (7) | 20 (12) |
| >28 d | 127 (32) | 35 (16) | 92 (53) |
Abbreviations: EDH, epidural hemorrhage; GCS, Glasgow Outcome Scale; PTCS, posttraumatic confusional state; SAH, subarachnoid hemorrhage; SDH, subdural hemorrhage; TBI, traumatic brain injury.
Seventeen patients had missing information on comorbidity category.
Work/education status was assessed at 6-month follow-up. Fourteen patients had missing work/education status.
Eight patients had unknown cause of injury.
In 17 patients, it was not possible to determine a reliable GCS score at admission.
Six patients died because of other reasons before the follow-up at 12 months.
PTCS duration and 12-month functional outcome in moderate and severe TBI
The median GOSE score at 12 months was 6 (IQR, 3) in the total sample; 7 for moderate TBI, and 5 for severe TBI (see table 1). Figure 2 and tables 2a (moderate TBI) and 2b (severe TBI) show that the GOSE score decreased with increasing PTCS duration (moderate TBI: Spearman's ρ -0.371, P<.001, severe TBI: Spearman's ρ -0.664, P<.001). When PTCS lasted 0-7 days, a good recovery (GOSE score 7-8) was observed in 78% of moderate TBI cases and 71% of severe TBI cases. When PTCS lasted <28 days, a GOSE score of 1-4 was observed in only 3% of cases.
Fig 2.
Boxplots showing the distribution of the GOSE in relation to PTCS duration.
Table 2a.
The general function (assessed with the GOSE) at 12 months postinjury in relation to the duration of PTCS in patients with moderate TBI.
| PTCS Duration, n (%) |
|||||
|---|---|---|---|---|---|
| 0-7 days | 8-14 days | 15-21 days | 22-28 days | >28 days | |
| Days | n=130 | n=28* | n=14 | n=15 | n=35† |
| GOSE Score | |||||
| Median (IQR) | 8 (7, 8) | 7 (6, 8) | 6 (5, 8) | 6 (5, 6) | 5 (3, 8) |
| 8 | 72 (55) | 10 (40) | 4 (29) | 2 (13) | 6 (19) |
| 7 | 30 (23) | 6 (24) | 2 (14) | 1 (7) | 3 (9) |
| 6 | 18 (14) | 3 (12) | 4 (29) | 5 (33) | 4 (13) |
| 5 | 7 (5) | 4 (16) | 3 (21) | 5 (33) | 3 (9) |
| 4 | 1 (1) | 1 (4) | 0 | 2 (13) | 2 (6) |
| 3 | 2 (2) | 1 (4) | 1 (7) | 0 | 13 (41) |
| 2 | 0 | 0 | 0 | 0 | 0 |
| 1 | 0 | 0 | 0 | 0 | 1 (3) |
Abbreviations: GOSE, Glasgow Outcome Scale–Extended; PTCS, posttraumatic confusional state.
Deaths because of other cause were not included in the percentages. The percentages are calculated from 25 patients.
Deaths because of other cause were not included in the percentages. The percentages are calculated from 32 patients.
Table 2b.
The general function (assessed with the GOSE) at 12 months postinjury in relation to the duration of PTCS in patients with severe TBI.
| PTCS Duration, n (%) |
|||||
|---|---|---|---|---|---|
| 0-7 days | 8-14 days | 15-21 days | 22-28 days | >28 days | |
| Days | n=31 | n=19 | n=11 | n=20 | n=92 |
| GOSE Score | |||||
| Median (IQR) | 7 (6, 8) | 7 (6, 8) | 7 (6, 8) | 6 (5, 6) | 4 (3, 5) |
| 8 | 15 (48) | 6 (32) | 4 (36) | 2 (10) | 4 (4) |
| 7 | 7 (23) | 6 (32) | 2 (18) | 3 (15) | 2 (2) |
| 6 | 3 (10) | 3 (16) | 4 (36) | 9 (45) | 13 (14) |
| 5 | 6 (19) | 4 (21) | 1 (9) | 5 (25) | 23 (25) |
| 4 | 0 | 0 | 0 | 0 | 12 (13) |
| 3 | 0 | 0 | 0 | 1 (5) | 29 (31) |
| 2 | 0 | 0 | 0 | 0 | 2 (2) |
| 1 | 0 | 0 | 0 | 0 | 7 (8) |
Abbreviations: GOSE, Glasgow Outcome Scale–Extended; PTCS, posttraumatic confusional state.
Prediction of outcome in moderate TBI
Longer PTCS duration was a significant predictor of worse outcome in most multivariable models, which included age, sex, the GCS score, pupillary dilatation, and the Rotterdam CT score (tables 3a and 4a). The odds ratios (ORs) for PTCS duration mostly increased with longer PTCS duration, except for lower ORs in the >28 days’ category than in the 15-28 days when the worse outcome was defined as a GOSE score of ≤7 and ≤6. The effects of age and the GCS score were slightly reduced in all multivariable models when PTCS duration was added, as was the effect of the Rotterdam CT score in most models. Including PTCS in the multivariable models mostly improved model fit, according to model performance measures (eg, Nagelkerke R2, AIC, and BIC, tables 3a, 4a, and 5a). Results from the univariable logistic regression analyses are shown in Supplemental Table S3 (available online only at http://www.archives-pmr.org/).
Table 3a.
Prediction of GOSE score ≤6 in moderate TBI 12 months after injury (only multivariable analyses).*
| Multivariable Analysis Without PTCS Duration†,‡ |
Multivariable Analysis With PTCS Duration‡,§,|| |
|||||||
|---|---|---|---|---|---|---|---|---|
| n | OR | 95% CI | P Value | n | OR | 95% CI | P Value | |
| Age | 195 | 1.03 | 1.01-1.05 | .004 | 195 | 1.02 | 1.00-1.04 | .065 |
| Sex | ||||||||
| Male | 135 | Ref. | . | . | 135 | Ref. | . | . |
| Female | 60 | 2.10 | 1.02-4.32 | .045 | 60 | 2.58 | 1.19-5.58 | .016 |
| GCS score | 195 | 0.70 | 0.55-0.88 | .003 | 195 | 0.77 | 0.60-1.00 | .051 |
| Pupillary dilatation | ||||||||
| No dilatation | 184 | Ref. | . | . | 184 | Ref. | . | . |
| Dilatation | 11 | 1.30 | 0.32-5.25 | .709 | 11 | 1.00 | 0.20-4.94 | .999 |
| Rotterdam CT score | .004 | .035 | ||||||
| 1 and 2 | 67 | Ref. | . | . | 67 | Ref. | . | . |
| 3 | 94 | 1.60 | 0.71-3.61 | .261 | 94 | 1.37 | 0.58-3.25 | .477 |
| 4, 5, and 6 | 34 | 5.30 | 1.90-14.76 | .001 | 34 | 3.93 | 1.32-11.70 | .014 |
| PTCS duration | .001 | |||||||
| 0-7 d | . | . | . | . | 117 | Ref. | . | . |
| 8-14 d | . | . | . | . | 23 | 1.14 | 0.38-3.46 | .811 |
| 15-28 d | . | . | . | . | 27 | 5.90 | 2.15-16.20 | <.001 |
| >28 d | . | . | . | . | 28 | 4.27 | 1.50-12.16 | .006 |
| Nagelkerke R2 | 0.260 | 0.353 | ||||||
| AIC | 225 | 215 | ||||||
| BIC | 248 | 247 | ||||||
Abbreviations: CI, confidence interval; CT, computed tomography; GCS, Glasgow Coma Scale; GOSE, Glasgow Outcome Scale–Extended; OR, odds ratio; PTCS, posttraumatic confusional state.
Six patients died because of other causes within 1 year and were therefore not included in the analyses.
Hosmer and Lemeshow test, P=.996.
Twenty-one patients were excluded from the analysis because of missing or no reliable GCS score and/or missing information on pupillary dilatation.
Hosmer and Lemeshow test, P=.492.
The multivariable model including PTCS duration had a significant improvement in fit (P value <.001) compared with the multivariable model without PTCS duration, according to the likelihood ratio chi-square test.
Table 4a.
Prediction of GOSE score ≤5 in moderate TBI 12 months after injury (only multivariable analyses).*
| Multivariable Analysis Without PTCS Duration†,‡ |
Multivariable Analysis With PTCS Duration‡,§,|| |
|||||||
|---|---|---|---|---|---|---|---|---|
| n | OR | 95% CI | P Value | n | OR | 95% CI | P Value | |
| Age | 195 | 1.04 | 1.02-1.07 | .001 | 195 | 1.03 | 1.00-1.06 | .046 |
| Sex | ||||||||
| Male | 135 | Ref. | . | . | 135 | Ref. | . | . |
| Female | 60 | 1.37 | 0.58-3.26 | .475 | 60 | 1.73 | 0.67-4.43 | .255 |
| GCS score | 195 | 0.71 | 0.54-0.95 | .018 | 195 | 0.83 | 0.61-1.13 | .240 |
| Pupillary dilatation | ||||||||
| No dilatation | 184 | Ref. | . | . | 184 | Ref. | . | . |
| Dilatation | 11 | 1.48 | 0.35-6.26 | .592 | 11 | 0.98 | 0.19-4.95 | .980 |
| Rotterdam CT score | .004 | .048 | ||||||
| 1 and 2 | 67 | Ref. | . | . | 67 | Ref. | . | . |
| 3 | 94 | 1.92 | 0.62-5.95 | .261 | 94 | 1.57 | 0.46-5.31 | .473 |
| 4, 5, and 6 | 34 | 6.77 | 1.96-23.43 | .003 | 34 | 4.44 | 1.16-16.99 | .029 |
| PTCS duration | .001 | |||||||
| 0-7 d | . | . | . | . | 117 | Ref. | . | . |
| 8-14 d | . | . | . | . | 23 | 1.86 | 0.46-7.53 | .386 |
| 15-28 d | . | . | . | . | 27 | 5.52 | 1.71-17.85 | .004 |
| >28 d | . | . | . | . | 28 | 8.44 | 2.67-26.65 | <.001 |
| Nagelkerke R2 | 0.279 | 0.386 | ||||||
| AIC | 169 | 158 | ||||||
| BIC | 192 | 191 | ||||||
Abbreviations: CI, confidence interval; CT, computed tomography; GCS, Glasgow Coma Scale; GOSE, Glasgow Outcome Scale–Extended; OR, odds ratio; PTCS, posttraumatic confusional state.
Six patients died because of other causes within 1 year and were therefore not included in the analyses.
Hosmer and Lemeshow test, P=.345.
Twenty-one patients were excluded from the analysis because of missing or no reliable GCS score and/or missing information on pupillary dilatation.
Hosmer and Lemeshow test, P=.879.
The multivariable model including PTCS duration had a significant improvement in fit (P value <.001) compared with the multivariable model without PTCS duration, according to the likelihood ratio chi-square test.
Table 5a.
Prediction of GOSE score ≤7 in moderate TBI 12 months after injury (only multivariable analyses).*
| Multivariable Analysis Without PTCS Duration†,‡ |
Multivariable Analysis With PTCS Duration‡,§,|| |
|||||||
|---|---|---|---|---|---|---|---|---|
| n | OR | 95% CI | P Value | n | OR | 95% CI | P Value | |
| Age | 195 | 1.03 | 1.01-1.04 | .004 | 195 | 1.02 | 1.00-1.04 | .020 |
| Sex | ||||||||
| Male | 135 | Ref. | . | . | 135 | Ref. | . | . |
| Female | 60 | 2.17 | 1.06-4.44 | .034 | 60 | 2.36 | 1.14-4.88 | .021 |
| GCS score | 195 | 0.78 | 0.62-0.98 | .035 | 195 | 0.83 | 0.65-1.06 | .135 |
| Pupillary dilatation | ||||||||
| No dilatation | 184 | Ref. | . | . | 184 | Ref. | . | . |
| Dilatation | 11 | 2.31 | 0.45-11.96 | .319 | 11 | 2.26 | 0.41-12.46 | .349 |
| Rotterdam CT score | .037 | .117 | ||||||
| 1 and 2 | 67 | Ref. | . | . | 67 | Ref. | . | . |
| 3 | 94 | 1.92 | 0.94-3.95 | .075 | 94 | 1.78 | 0.86-3.69 | .123 |
| 4, 5, and 6 | 34 | 3.52 | 1.30-9.56 | .014 | 34 | 2.79 | 0.99-7.89 | .052 |
| PTCS duration | .125 | |||||||
| 0-7 d | . | . | . | . | 117 | Ref. | . | . |
| 8-14 d | . | . | . | . | 23 | 1.25 | 0.46-3.39 | .656 |
| 15-28 d | . | . | . | . | 27 | 3.11 | 1.08-8.95 | .035 |
| >28 d | . | . | . | . | 28 | 2.31 | 0.77-6.91 | .136 |
| Nagelkerke R2 | 0.222 | 0.256 | ||||||
| AIC | 247 | 247 | ||||||
| BIC | 270 | 279 | ||||||
Abbreviations: CI, confidence interval; CT, computed tomography; GCS, Glasgow Coma Scale; GOSE, Glasgow Outcome Scale–Extended; OR, odds ratio; PTCS, posttraumatic confusional state.
Six patients died because of other causes within 1 year and were therefore not included in the analyses.
Hosmer and Lemeshow test, P=.898.
Twenty-one patients were excluded from the analysis because of missing or no reliable GCS score and/or missing information on pupillary dilatation.
Hosmer and Lemeshow test, P=.295.
The multivariable model including PTCS duration did not improve the fit (P value.11) compared with the multivariable model without PTCS duration, according to the likelihood ratio chi-square test.
Positive and negative predictive values of PTCS duration >28 days were 50% and 96%, respectively, when predicting GOSE score ≤4 (table 6a).
Table 6a.
The sensitivity and specificity of the PTCS duration on the GOSE score ≤4 in moderate TBI.
| PTCS Duration ↓ | GOSE |
||
|---|---|---|---|
| Score ≤4, n | Score >4, n | ||
| >28 d | 16 | 16 | PPV: 50% |
| 0-28 d | 8 | 176 | NPV: 96% |
| Sensitivity: 67% | Specificity: 92% | ||
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; PTCS, posttraumatic confusional state.
Prediction of outcome in severe TBI
Longer PTCS duration was the only significant predictor of worse outcome in the multivariable models (tables 3b, 4b, and 5b). The ORs for worse outcome increased with longer PTCS duration. The effects of the GCS score and the Rotterdam CT score were reduced in all multivariable models that included PTCS duration as a predictor. When the cutoff was set at a GOSE score of ≤5, the effect of pupillary dilatation was also reduced. Including PTCS in the multivariable models improved model fit, according to model performance measures (eg, Nagelkerke R2, AIC, and BIC, table 3b, 4b, and 5b). Results from the univariable logistic regression analyses are shown in Supplemental Table S4 (available online only at http://www.archives-pmr.org/)..
Table 3b.
Prediction of GOSE score ≤6 in severe TBI 12 months after injury (only multivariable analyses).
| Multivariable Analysis Without PTCS Duration*,† |
Multivariable Analysis With PTCS Duration†,‡,§ |
|||||||
|---|---|---|---|---|---|---|---|---|
| n | OR | 95% CI | P Value | n | OR | 95% CI | P Value | |
| Age | 167 | 1.02 | 1.00-1.04 | .128 | 167 | 1.01 | 0.99-1.04 | .347 |
| Sex | ||||||||
| Male | 134 | Ref. | . | . | 134 | Ref. | . | . |
| Female | 33 | 1.48 | 0.54-4.08 | .446 | 33 | 2.24 | 0.71-7.08 | .168 |
| GCS score | 167 | 0.77 | 0.62-0.96 | .019 | 167 | 0.95 | 0.73-1.24 | .724 |
| Pupillary dilatation | ||||||||
| No dilatation | 125 | Ref. | . | . | 125 | Ref. | . | . |
| Dilatation | 42 | 1.50 | 0.56-4.05 | .423 | 42 | 0.76 | 0.24-2.46 | .646 |
| Rotterdam CT score | .005 | .223 | ||||||
| 1 and 2 | 43 | Ref. | . | . | 43 | Ref. | . | . |
| 3 | 60 | 1.88 | 0.78-4.56 | .160 | 60 | 1.13 | 0.40-3.24 | .816 |
| 4, 5, and 6 | 64 | 5.77 | 2.01-16.60 | .001 | 64 | 2.68 | 0.82-8.69 | .101 |
| PTCS duration | <.001 | |||||||
| 0-7 d | . | . | . | . | 31 | Ref. | . | . |
| 8-14 d | . | . | . | . | 18 | 1.54 | 0.43-5.53 | 0.511 |
| 15-28 d | . | . | . | . | 31 | 4.02 | 1.29-12.57 | 0.017 |
| >28 d | . | . | . | . | 87 | 25.74 | 6.99-94.83 | <0.001 |
| Nagelkerke R2 | 0.242 | 0.458 | ||||||
| AIC | 185 | 157 | ||||||
| BIC | 207 | 189 | ||||||
Abbreviations: CI, confidence interval; CT, computed tomography; GCS, Glasgow Coma Scale; GOSE, Glasgow Outcome Scale–Extended; OR, odds ratio; PTCS, posttraumatic confusional state.
Hosmer and Lemeshow test, P=.154.
Six patients were excluded from the analysis because of missing or no reliable GCS score and/or missing information on pupillary dilatation.
Hosmer and Lemeshow test, P=.958.
The multivariable model including PTCS duration had a significant improvement in fit (P value <.001) compared with the multivariable model without PTCS duration, according to the likelihood ratio chi-square test.
Table 4b.
Prediction of GOSE score ≤5 in severe TBI 12 months after injury (only multivariable analyses).
| Multivariable Analysis Without PTCS Duration*,† |
Multivariable Analysis With PTCS Duration†,‡,§ |
|||||||
|---|---|---|---|---|---|---|---|---|
| n | OR | 95% CI | P value | n | OR | 95% CI | P value | |
| Age | 167 | 1.02 | 1.00-1.04 | .080 | 167 | 1.02 | 0.99-1.04 | .199 |
| Sex | ||||||||
| Male | 134 | Ref. | . | . | 134 | Ref. | . | . |
| Female | 33 | 0.60 | 0.24-1.47 | .262 | 33 | 0.56 | 0.20-1.56 | .269 |
| GCS score | 167 | 0.79 | 0.64-0.96 | .020 | 167 | 0.92 | 0.73-1.16 | .479 |
| Pupillary dilatation | ||||||||
| No dilatation | 125 | Ref. | . | . | 125 | Ref. | . | . |
| Dilatation | 42 | 2.65 | 1.14-6.17 | .024 | 42 | 1.91 | 0.74-4.93 | .183 |
| Rotterdam CT score | .009 | .431 | ||||||
| 1 and 2 | 43 | Ref. | . | . | 43 | Ref. | . | . |
| 3 | 60 | 2.92 | 1.14-7.51 | .026 | 60 | 1.64 | 0.57-4.73 | .360 |
| 4, 5, and 6 | 64 | 4.32 | 1.68-11.07 | .002 | 64 | 2.02 | 0.69-5.86 | .198 |
| PTCS duration | <.001 | |||||||
| 0-7 d | . | . | . | . | 31 | Ref. | . | . |
| 8-14 d | . | . | . | . | 18 | 1.09 | 0.25-4.74 | .907 |
| 15-28 d | . | . | . | . | 31 | 0.92 | 0.25-3.38 | .905 |
| >28 d | . | . | . | . | 87 | 8.64 | 2.76-27.06 | <.001 |
| Nagelkerke R2 | 0.250 | 0.428 | ||||||
| AIC | 209 | 186 | ||||||
| BIC | 231 | 217 | ||||||
Abbreviations: CI, confidence interval; CT, computed tomography; GCS, Glasgow Coma Scale; GOSE, Glasgow Outcome Scale–Extended; OR, odds ratio; PTCS, posttraumatic confusional state.
Hosmer and Lemeshow test, P=.838.
Six patients were excluded from the analysis because of missing or no reliable GCS score and/or missing information on pupillary dilatation.
Hosmer and Lemeshow test, P=.823.
The multivariable model including PTCS duration had a significant improvement in fit (P value <.001) compared with the multivariable model without PTCS duration, according to the likelihood ratio chi-square test.
Table 5b.
Prediction of GOSE score ≤7 in severe TBI 12 months after injury (only multivariable analyses).
| Multivariable Analysis Without PTCS Duration*,† |
Multivariable Analysis With PTCS Duration†,‡,§ |
|||||||
|---|---|---|---|---|---|---|---|---|
| n | OR | 95% CI | P Value | n | OR | 95% CI | P Value | |
| Age | 167 | 1.02 | 0.99-1.05 | .155 | 167 | 1.01 | 0.99-1.04 | .347 |
| Sex | ||||||||
| Male | 134 | Ref. | . | . | 134 | Ref. | . | . |
| Female | 33 | 1.45 | 0.46-4.58 | .525 | 33 | 1.92 | 0.57-6.49 | .296 |
| GCS score | 167 | 0.75 | 0.58-0.97 | .028 | 167 | 0.90 | 0.67-1.20 | .452 |
| Pupillary dilatation | ||||||||
| No dilatation | 125 | Ref. | . | . | 125 | Ref. | . | . |
| Dilatation | 42 | 1.03 | 0.33-3.22 | .957 | 42 | 0.56 | 0.16-2.04 | .383 |
| Rotterdam CT score | .071 | .439 | ||||||
| 1 and 2 | 43 | Ref. | . | . | 43 | Ref. | . | . |
| 3 | 60 | 1.31 | 0.50-3.42 | .588 | 60 | 0.88 | 0.30-2.60 | .811 |
| 4, 5, and 6 | 64 | 4.42 | 1.22-16.06 | .024 | 64 | 2.11 | 0.53-8.44 | .290 |
| PTCS duration | .002 | |||||||
| 0-7 d | . | . | . | 31 | Ref. | . | . | |
| 8-14 d | . | . | . | 18 | 1.79 | 0.51-6.24 | .364 | |
| 15-28 d | . | . | . | 31 | 3.33 | 1.01-11.02 | .049 | |
| >28 d | . | . | . | 87 | 14.94 | 3.59-62.25 | <.001 | |
| Nagelkerke R2 | 0.175 | 0.317 | ||||||
| AIC | 154 | 143 | ||||||
| BIC | 176 | 174 | ||||||
Abbreviations: CI, confidence interval; CT, computed tomography; GCS, Glasgow Coma Scale; GOSE, Glasgow Outcome Scale–Extended; OR, odds ratio; PTCS, posttraumatic confusional state.
Hosmer and Lemeshow test, P=.658.
Six patients were excluded from the analysis because of missing or no reliable GCS score and/or missing information on pupillary dilatation.
Hosmer and Lemeshow test, P=.490.
The multivariable model including PTCS duration had a significant improvement in fit (P value <.001) compared with the multivariable model without PTCS duration, according to the likelihood ratio chi-square test.
Positive and negative predictive values of PTCS duration >28 days were 54% and 99%, respectively, when predicting GOSE score ≤4 (table 6b).
Table 6b.
The sensitivity and specificity of the PTCS duration on the GOSE score ≤4 in severe TBI.
| PTCS Duration ↓ | GOSE |
||
|---|---|---|---|
| Score ≤4, n | Score >4, n | ||
| >28 d | 50 | 42 | PPV: 54% |
| 0-28 d | 1 | 80 | NPV: 99% |
| Sensitivity: 98% | Specificity: 66% | ||
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; PTCS, posttraumatic confusional state.
Discussion
In this large cohort study of unselected patients in the acute stage, longer PTCS duration was a strong predictor of worse outcome. Notably, in severe TBI, PTCS duration outperformed all other predictors when added to the multivariable models. In this neurosurgical cohort, almost none of the patients with PTCS duration ≤28 days had severe disability (GOSE score ≤4) at 12 months postinjury, resulting in negative predictive values of 96% and 99% for PTCS duration >28 days in moderate and severe TBI, respectively.
A strong association between PTCS duration and outcome was expected. Several previous studies have found longer PTCS duration to be associated with worse outcome or worse cognitive ability in severe TBI.9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,36 This study adds to the existing literature in 2 ways. First, it demonstrates this association in a large unselected neurosurgical cohort using available clinical descriptions, showing that PTCS estimation is feasible in a hospital setting. Second, it analyzes PTCS duration separately in moderate and severe TBI.
In moderate TBI, although longer PTCS duration was a significant predictor of worse outcome (GOSE scores ≤6 and ≤5), it did not improve the models as much as it did for severe TBI. PTCS duration may not explain as much of the variation in outcome in moderate TBI because many patients achieve GOSE scores in the upper range, and other factors may be more important. It has been suggested that a range of psychosocial factors are important for outcome in moderate TBI,34 as such factors have been shown to predict outcome in mild TBI.37
In severe TBI, PTCS duration was a significant predictor of worse outcome in the multivariable models, and adding PTCS duration improved model performance. Neither the GCS score nor the worst Rotterdam CT score was a significant predictor of outcome in severe TBI when PTCS duration was included in the models. Previous studies have questioned the predictive value of these injury-related variables.9,25 Because PTCS duration is assessed at a later time point than the GCS score, it might not be surprising that PTCS duration is a stronger predictor of outcome. The GCS score is also influenced by factors such as intoxication, sedation, and severe extracranial injuries.38 Moreover, patients with similar CT findings may have different PTCS durations because of co-occurring traumatic axonal injuries (TAIs) in the brainstem, basal ganglia, or thalamus.39 Notably, patients with TAI often have prolonged PTCS duration39 and a high risk of poor outcome.40 Therefore, using only the acute GCS score and CT-based information to predict outcome after TBI may result in the loss of important information.
An interesting finding was that experiencing <1 week of PTCS implied a high probability of good recovery (GOSE score 7-8), regardless of whether the TBI was classified as moderate or severe based on the GCS score at admission. This clinically useful observation aligns with a recent systematic review and meta-analysis where PTCS duration <7 days was a reliable predictor of good recovery.36 Hence, in the hospital setting, the PTCS duration of 0-7 days could enhance future classification systems for TBI and serve as a clinical modifier beyond the GCS score.
Another striking finding was that almost all patients with PTCS duration ≤28 days (96% of moderate and 99% of severe TBI) were independent in daily life (GOSE score ≥5) within 12 months postinjury. The CRASH and IMPACT models are the most used and validated prognostic models of mortality or dependency after moderate and severe TBI.2 Despite their widespread use, the predictors in CRASH41 and IMPACT42 fail to explain much of the variation in outcome.2 As a result, the research community calls for new additional or dynamic predictors.43 Our findings underscore the potential of PTCS duration as a predictor of independence in survivors of moderate and severe TBI.
Based on the results of this study, we believe that frequent assessment and description of the clinical characteristics of PTCS can aid clinicians in evaluating TBI severity in individual patients and improve research. Validated measures of PTCS are recommended when possible. However, our study shows that more pragmatic, qualitatively based estimates of PTCS duration are of great value. Such estimates are more feasible to collect in busy acute hospital wards. Pragmatic assessment of PTCS duration could be implemented in hospital routines and large-scale studies on TBI without high costs.
Study limitations
Strengths of this study include the large sample size, high inclusion and follow-up rates, and the inclusion of moderate TBI. However, the study has some limitations. First, in a large proportion of cases, PTCS duration was estimated retrospectively by reviewing medical records, which may be less accurate. Factors such as anesthesia required for surgery, analgesics, and the onset of acute or posttraumatic stress disorder might influence the patient´s consciousness in addition to the TBI.44 When factors likely to influence PTCS duration were known, the duration was considered unreliable. Second, PTCS duration was unknown in 11% (n=49) of patients with follow-up because of the lack of information in medical records or transfer to hospitals outside of the health region. Third, by estimating PTCS duration into week intervals (and thereby using it as a categorical variable in the analyses), the statistical models were less refined compared with using a more fine-graded measure. Fourth, analyzing several GOSE cutoffs may increase the risk of type I error. Fifth, patients were included from a regional neurosurgical referral facility, and findings may not be representative of patients in other regions or countries, or of patients solely treated at general hospitals. In a previous study, we showed that patients with low risk of deterioration and elderly patients (mostly ≥80y) were less likely to be transferred to the regional trauma center.45 Therefore, external validation of our findings is needed.
Conclusions
Longer PTCS duration was a strong predictor of worse outcome 12 months after the injury, especially in severe TBI. Almost all patients with moderate and severe TBI with PTCS duration ≤28 days were independent in daily activities, and most patients with PTCS duration 0-7 days had a good recovery. Including a pragmatically estimated PTCS duration in the predictive models improved their performance, particularly in severe TBI. The strong predictive value of PTCS duration in this neurosurgical cohort demonstrates that PTCS duration could be used to modify the severity classification of TBI and predict outcome. Based on our results, we recommend that clinical characteristics of PTCS are continuously assessed and PTCS duration is documented in both hospital and rehabilitation settings.
Disclosure
The investigators have no financial or nonfinancial disclosures to make in relation to this project.
Acknowledgments
Acknowledgments
We would like to express our gratitude to Beate Mærk Voll, Stine Borgen Lund, Otto Noralf Aarhaug, and Torun Elsie Farnes, for the management of the study database in the Trondheim moderate and severe TBI study. We further thank residents at the Department of Neurosurgery for their contribution to the inclusion of patients in the Trondheim moderate and severe TBI study.
Declaration of Generative AI and AI-Assisted Technologies in the Writing Process
During the preparation of this work, the authors used Copilot for proofread and to improve readability. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
Footnotes
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.arrct.2025.100446.
Appendix. Supplementary materials
References
- 1.Maas A.I., Lingsma H.F., Roozenbek B. In: Predicting outcome after traumatic brain injury. Grafman J., Salazar A.M., editors. Elsevier; Amsterdam, Netherlands: 2015. Handbook of clinical neurology; pp. 455–474. [DOI] [PubMed] [Google Scholar]
- 2.Dijkland S.A., Foks K.A., Polinder S., et al. Prognosis in moderate and severe traumatic brain injury: a systematic review of contemporary models and validation studies. J Neurotrauma. 2020;37:1–13. doi: 10.1089/neu.2019.6401. [DOI] [PubMed] [Google Scholar]
- 3.Lingsma H.F., Roozenbeek B., Steyerberg E.W., Murray G.D., Maas AI. Early prognosis in traumatic brain injury: from prophecies to predictions. Lancet Neurol. 2010;9:543–554. doi: 10.1016/S1474-4422(10)70065-X. [DOI] [PubMed] [Google Scholar]
- 4.Sherer M., Katz D.I., Bodien Y.G., et al. Post-traumatic confusional state: a case definition and diagnostic criteria. Arch Phys Med Rehabil. 2020;101:2041–2050. doi: 10.1016/j.apmr.2020.06.021. [DOI] [PubMed] [Google Scholar]
- 5.Ponsford J., Carrier S., Hicks A., McKay A. Assessment and management of patients in the acute stages of recovery after traumatic brain injury in adults: a worldwide survey. J Neurotrauma. 2021;38:1060–1067. doi: 10.1089/neu.2020.7299. [DOI] [PubMed] [Google Scholar]
- 6.Russell W.R., Nathan PW. Traumatic amnesia. Brain. 1946;69:280–300. doi: 10.1093/brain/69.4.280. [DOI] [PubMed] [Google Scholar]
- 7.Symonds C.P., Russell WR. Accidental head injuries. Lancet. 1943;241:7–10. [Google Scholar]
- 8.Stuss D.T., Binns M.A., Carruth F.G., et al. The acute period of recovery from traumatic brain injury: posttraumatic amnesia or posttraumatic confusional state? J Neurosurg. 1999;90:635–643. doi: 10.3171/jns.1999.90.4.0635. [DOI] [PubMed] [Google Scholar]
- 9.Sveen U., Røe C., Sigurdardottir S., et al. Rehabilitation pathways and functional independence one year after severe traumatic brain injury. Eur J Phys Rehabil Med. 2016;52:650–661. [PubMed] [Google Scholar]
- 10.Sommer J.B., Norup A., Poulsen I., Morgensen J. Cognitive activity limitations one year post-trauma in patients admitted to sub-acute rehabilitation after severe traumatic brain injury. J Rehabil Med. 2013;45:778–784. doi: 10.2340/16501977-1208. [DOI] [PubMed] [Google Scholar]
- 11.Haslam C., Batchelor J., Fearnside M.R., Haslam S.A., Hawkins S., Kenway E. Post-coma disturbance and post-traumatic amnesia as nonlinear predictors of cognitive outcome following severe closed head injury: findings from the Westmead Head Injury Project. Brain Inj. 1994;8:519–528. doi: 10.3109/02699059409151004. [DOI] [PubMed] [Google Scholar]
- 12.Quach N.T., Ehsanian R., Dirlikov B., et al. Burden of care implications and association of intracranial hypertension with extremely severe post-traumatic amnesia after traumatic brain injury: a 5-year retrospective longitudinal study. Front Neurol. 2019;10:34. doi: 10.3389/fneur.2019.00034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.van der Naalt J., van Zomeren A.H., Sluiter W.J., Minderhoud JM. One year outcome in mild to moderate head injury: the predictive value of acute injury characteristics related to complaints and return to work. J Neurol Neurosurg Psychiatry. 1999;66:207–213. doi: 10.1136/jnnp.66.2.207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sherer M., Yablon S.A., Nakase-Richardson R., Nick TG. Effect of severity of post-traumatic confusion and its constituent symptoms on outcome after traumatic brain injury. Arch Phys Med Rehabil. 2008;89:42–47. doi: 10.1016/j.apmr.2007.08.128. [DOI] [PubMed] [Google Scholar]
- 15.Walker W.C., Ketchum J.M., Marwitz J.H., et al. A multicentre study on the clinical utility of post-traumatic amnesia duration in predicting global outcome after moderate-severe traumatic brain injury. J Neurol Neurosurg Psychiatry. 2010;81:87–89. doi: 10.1136/jnnp.2008.161570. [DOI] [PubMed] [Google Scholar]
- 16.Brown A.W., Malec J.F., McClelland R.L., Diehl N.N., Englander J., Cifu DX. Clinical elements that predict outcome after traumatic brain injury: a prospective multicenter recursive partitioning (decision-tree) analysis. J Neurotrauma. 2005;22:1040–1051. doi: 10.1089/neu.2005.22.1040. [DOI] [PubMed] [Google Scholar]
- 17.Nakase-Richardson R., Sepehri A., Sherer M., Yablon S.A., Evans C., Mani T. Classification schema of posttraumatic amnesia duration-based injury severity relative to 1-year outcome: analysis of individuals with moderate and severe traumatic brain injury. Arch Phys Med Rehabil. 2009;90:17–19. doi: 10.1016/j.apmr.2008.06.030. [DOI] [PubMed] [Google Scholar]
- 18.Gardner R.C., Cheng J., Ferguson A.R., et al. Divergent six month functional recovery trajectories and predictors after traumatic brain injury: novel insights from the citicoline brain injury treatment trial study. J Neurotrauma. 2019;36:2521–2532. doi: 10.1089/neu.2018.6167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Aldossary N.M., Kotb M.A., Kamal AM. Predictive value of early MRI findings on neurocognitive and psychiatric outcomes in patients with severe traumatic brain injury. J Affect Disord. 2019;243:1–7. doi: 10.1016/j.jad.2018.09.001. [DOI] [PubMed] [Google Scholar]
- 20.Avesani R., Salvi L., Rigoli G., Gambini MG. Reintegration after severe brain injury: a retrospective study. Brain Inj. 2005;19:933–939. doi: 10.1080/02699050400002371. [DOI] [PubMed] [Google Scholar]
- 21.Dahdah M.N., Hofmann M., Pretz C., et al. Factors associated with rehabilitation outcomes after traumatic brain injury: comparing functional outcomes between TBIMS centers using hierarchical linear modeling. J Head Trauma Rehabil. 2017;32:E1–10. doi: 10.1097/HTR.0000000000000270. [DOI] [PubMed] [Google Scholar]
- 22.Eastvold A.D., Walker W.C., Curtiss G., Schwab K., Vanderploeg RD. The differential contributions of posttraumatic amnesia duration and time since injury in prediction of functional outcomes following moderate-to-severe traumatic brain injury. J Head Trauma Rehabil. 2013;28:48–58. doi: 10.1097/HTR.0b013e31823c9317. [DOI] [PubMed] [Google Scholar]
- 23.Ketchum J.M., Almaz Getachew M., Krch D., et al. Early predictors of employment outcomes 1 year post traumatic brain injury in a population of Hispanic individuals. NeuroRehabilitation. 2012;30:13–22. doi: 10.3233/NRE-2011-0723. [DOI] [PubMed] [Google Scholar]
- 24.Ponsford J., Draper K., Schönberger M. Functional outcome 10 years after traumatic brain injury: its relationship with demographic, injury severity, and cognitive and emotional status. J Int Neuropsychol Soc. 2008;14:233–242. doi: 10.1017/S1355617708080272. [DOI] [PubMed] [Google Scholar]
- 25.Walker W.C., Stromberg K.A., Marwitz J.H., et al. Predicting long-term global outcome after traumatic brain injury: development of a practical prognostic tool using the traumatic brain injury model systems national database. J Neurotrauma. 2018;35:1587–1595. doi: 10.1089/neu.2017.5359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sigurdardottir S., Andelic N., Roe C., Schanke AK. Cognitive recovery and predictors of functional outcome 1 year after traumatic brain injury. J Int Neuropsychol Soc. 2009;15:740–750. doi: 10.1017/S1355617709990452. [DOI] [PubMed] [Google Scholar]
- 27.Bishara S.N., Partridge F.M., Godfrey H.P., Knight RG. Post-traumatic amnesia and Glasgow Coma Scale related to outcome in survivors in a consecutive series of patients with severe closed-head injury. Brain Inj. 1992;6:373–380. doi: 10.3109/02699059209034952. [DOI] [PubMed] [Google Scholar]
- 28.Bjarkø V.V., Skandsen T., Moen K.G., et al. Time of injury and relation to alcohol intoxication in moderate-to-severe traumatic brain injury: a decade-long prospective study. World Neurosurg. 2019;122:e684–e689. doi: 10.1016/j.wneu.2018.10.122. [DOI] [PubMed] [Google Scholar]
- 29.Maas A.I., Hukkelhoven C.W., Marshall L.F., Steyerberg EW. Prediction of outcome in traumatic brain injury with computed tomographic characteristics: a comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery. 2005;57:1173–1182. doi: 10.1227/01.neu.0000186013.63046.6b. [discussion 1173-82] [DOI] [PubMed] [Google Scholar]
- 30.Levin H.S., O'Donnell V.M., Grossman R.G. The galveston orientation and amnesia test. A practical scale to assess cognition after head injury. J Nerv Ment Dis. 1979;167:675–684. doi: 10.1097/00005053-197911000-00004. [DOI] [PubMed] [Google Scholar]
- 31.Jackson W.T., Novack T.A., Dowler RN. Effective serial measurement of cognitive orientation in rehabilitation: the Orientation Log. Arch Phys Med Rehabil. 1998;79:718–720. doi: 10.1016/s0003-9993(98)90051-x. [DOI] [PubMed] [Google Scholar]
- 32.Brown A.W., Malec J.F., Mandrekar J., et al. Predictive utility of weekly post-traumatic amnesia assessments after brain injury: a multicentre analysis. Brain Inj. 2010;24:472–478. doi: 10.3109/02699051003610466. [DOI] [PubMed] [Google Scholar]
- 33.Jennett B., Snoek J., Bond M.R., Brooks N. Disability after severe head injury: observations on the use of the Glasgow Outcome Scale. J Neurol Neurosurg Psychiatry. 1981;44:285–293. doi: 10.1136/jnnp.44.4.285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Einarsen C.E., van der Naalt J., Jacobs B., et al. Moderate traumatic brain injury: clinical characteristics and a prognostic model of 12-month outcome. World Neurosurg. 2018;114:e1199–e1210. doi: 10.1016/j.wneu.2018.03.176. [DOI] [PubMed] [Google Scholar]
- 35.Cremer O.L., Moons K.G., van Dijk G.W., van Balen P., Kalkman CJ. Prognosis following severe head injury: development and validation of a model for prediction of death, disability, and functional recovery. J Trauma. 2006;61:1484–1491. doi: 10.1097/01.ta.0000195981.63776.ba. [DOI] [PubMed] [Google Scholar]
- 36.Vile A.R., Jang K., Gourlay D., Marshman LAG. Posttraumatic amnesia: a systematic review and meta-analysis. Proposal for a new severity classification. World Neurosurg. 2022;162:e369–e393. doi: 10.1016/j.wneu.2022.03.018. [DOI] [PubMed] [Google Scholar]
- 37.Skandsen T., Stenberg J., Follestad T., et al. Personal factors associated with postconcussion symptoms 3 months after mild traumatic brain injury. Arch Phys Med Rehabil. 2021;102:1102–1112. doi: 10.1016/j.apmr.2020.10.106. [DOI] [PubMed] [Google Scholar]
- 38.Tenovuo O., Diaz-Arrastia R., Goldstein L.E., Sharp D.J., van der Naalt J., Zasler ND. Assessing the severity of traumatic brain injury-time for a change? J Clin Med. 2021;10:148. doi: 10.3390/jcm10010148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Moe H.K., Follestad T., Andelic N., et al. Traumatic axonal injury on clinical MRI: association with the Glasgow Coma Scale score at scene of injury or at admission and prolonged posttraumatic amnesia. J Neurosurg. 2020;135:562–573. doi: 10.3171/2020.6.JNS20112. [DOI] [PubMed] [Google Scholar]
- 40.Moen K.G., Brezova V., Skandsen T., Håberg A.K., Folvik M., Vik A. Traumatic axonal injury: the prognostic value of lesion load in corpus callosum, brain stem, and thalamus in different magnetic resonance imaging sequences. J Neurotrauma. 2014;31:1486–1496. doi: 10.1089/neu.2013.3258. [DOI] [PubMed] [Google Scholar]
- 41.Trial Collaborators MRC CRASH, Perel P., Arango M., et al. Predicting outcome after traumatic brain injury: practical prognostic models based on large cohort of international patients. BMJ. 2008;336:425–429. doi: 10.1136/bmj.39461.643438.25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Steyerberg E.W., Mushkudiani N., Perel P., et al. Predicting outcome after traumatic brain injury: development and international validation of prognostic scores based on admission characteristics. PLoS Med. 2008;5:e165. doi: 10.1371/journal.pmed.0050165. [discussion e165] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Maas A.I.R., Menon D.K., Manley G.T., et al. Traumatic brain injury: progress and challenges in prevention, clinical care, and research. Lancet Neurol. 2022;21:1004–1060. doi: 10.1016/S1474-4422(22)00309-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Friedland D., Swash M. Post-traumatic amnesia and confusional state: hazards of retrospective assessment. J Neurol Neurosurg Psychiatry. 2016;87:1068–1074. doi: 10.1136/jnnp-2015-312193. [DOI] [PubMed] [Google Scholar]
- 45.Rahim S., Laugsand E.A., Fyllingen E.H., et al. Moderate and severe traumatic brain injury in general hospitals: a ten-year population-based retrospective cohort study in central Norway. Scand J Trauma Resusc Emerg Med. 2022;30:68. doi: 10.1186/s13049-022-01050-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.