The increased risk of stroke in early insomnia following traumatic brain injury: a population-based cohort study.

OBJECTIVE: Insomnia, a common symptom after traumatic brain injury (TBI), may be a pre-symptom for developing stroke. This study aims to investigate whether insomnia is a potential risk factor for stroke after TBI, especially early insomnia. METHODS: Taiwan's Longitudinal Health Insurance Database 2000 from 1999 to 2013 was used in this cohort study. TBI patients with insomnia were selected based on the ICD-9-CM code (TBI: 801-804 and 850-854; insomnia: 307.4, 327, and 780.5). The outcome we were interested in was stroke (ICD-9-CM: 430-438). The incidence rate ratio of stroke between TBI with insomnia and the general population with insomnia was calculated by Poisson regression. The relative risk adjusted for potential confounding variables was estimated by Cox regression. RESULTS: For 1174 TBI patients with insomnia and 5870 general patients with insomnia, TBI patients have 209.85 incidence risk of new-onset stroke if they have insomnia. TBI patients have 2.28-fold (95% CI: 1.70-3.06) risk of new-onset stroke compared with the general population, even when controlling for age, gender, socioeconomic status, and comorbidities. The hazard ratio of new-onset stroke among different phases of new-onset insomnia after TBI surgery is 1.95-fold (95% CI: 1.05-3.62), 2.75-fold (95% CI: 1.73-4.37), and 2.66-fold (95% CI: 1.68-4.21) at ≤3, 3-12, and 12-24 months, compared with the general population with insomnia, respectively. CONCLUSION: TBI patients with insomnia have a higher risk of stroke compared with the general population with insomnia. Early new-onset insomnias after TBI will have higher risk of stroke. Therefore, we consider that insomnia could be a signal of the development of new-onset stroke in TBI patients.

The Association Between Ventriculo-Peritoneal Shunt and Acute Appendicitis in Patients with Traumatic Brain Injury: A 14-Year, Population-Based Study.

OBJECTIVE: The association between preexisting ventriculoperitoneal (VP) shunt and the risk of new-onset acute appendicitis in patients with traumatic brain injury (TBI) is not well established. The aim of the present study was to determine the relationships between VP shunt and acute appendicitis in patients with TBI. METHODS: A longitudinal cohort study matched by a propensity score in patients with TBI with (4781 patients) or without (9562 patients) VP shunt was conducted using the National Health Insurance Research Database in Taiwan between January 1993 and December 2013. RESULTS: The main outcome studied was diagnosis of acute appendicitis. The cumulative probability of acute appendicitis was not different between these 2 groups (P = 0.6244). A Cox model showed central nervous system (CNS) infection to be an independent predictor of acute appendicitis with an adjusted hazard ratio of 2.98. Patients with TBI with both a VP shunt and a CNS infection had a greater risk of developing new-onset acute appendicitis (hazard ratio 4.25; 95% confidence interval 1.84-9.81) compared patients with TBI without a VP shunt or CNS infection. CONCLUSIONS: We concluded that VP shunt is not a risk factor in the development of appendicitis in patients with TBI. Patients with TBI with a shunt and a CNS infection may have a greater risk of developing acute appendicitis. Therefore, care in avoiding CNS infection is a key for the prevention acute appendicitis in this patient population.

FOUR Score Predicts Early Outcome in Patients After Traumatic Brain Injury.

BACKGROUND: The aim of the study was to determine whether the Full Outline of UnResponsiveness (FOUR) score, which includes eyes opening (E), motor function (M), brainstem reflex (B), and respiratory pattern (R), can be used as an alternate method to the Glasgow Coma Scale (GCS) in predicting intensive care unit (ICU) mortality in traumatic brain injury (TBI) patients. METHODS: From January 2015 to June 2015, patients with isolated TBI admitted to the ICU were enrolled. Three advanced practice nurses administered the FOUR score, GCS, Acute Physiology and Chronic Health Evaluation II (APACHE II), and Therapeutic Intervention Scoring System (TISS) concurrently from ICU admissions. The endpoint of observation was mortality when the patients left the ICU. Data are presented as frequency with percentages, mean with standard deviation, or median with interquartile range. Each measurement tool used area under the receiver operating characteristic curve to compare the predictive power between these four tools. In addition, the difference between survival and death was estimated using the Wilcoxon rank sum test. RESULTS: From 55 TBI patients, males (72.73 %) were represented more than females, the mean age was 63.1 ± 17.9, and 19 of 55 observations (35 %) had a maximum FOUR score of 16. The overall mortality rate was 14.6 %. The area under the receiver operating characteristic curve was 74.47 % for the FOUR score, 74.73 % for the GCS, 81.78 % for the APACHE II, and 53.32 % for the TISS. The FOUR score has similar predictive power of mortality compared to the GCS and APACHE II. Each of the parameters-E, M, B, and R-of the FOUR score showed a significant difference between mortality and survival group, while the verbal and eye-opening components of the GCS did not. CONCLUSION: Having similar predictive power of mortality compared to the GCS and APACHE II, the FOUR score can be used as an alternative in the prediction of early mortality in TBI patients in the ICU.

Dynamic evolution of D-dimer level in cerebrospinal fluid predicts poor outcome in patients with spontaneous intracerebral hemorrhage combined with intraventricular hemorrhage.

The risk of mortality in patients with intracerebral hemorrhage (ICH) significantly increases when complicated by intraventricular hemorrhage (IVH). We hypothesize that serial measurement of cerebrospinal fluid (CSF) D-dimer levels in patients with both ICH and IVH may serve as an early marker of IVH severity. We performed a prospective study of 43 consecutive ICH patients combined with IVH and external ventricular drainage placement admitted in our institution from 2005-2006. IVH severity (Graeb score) and fibrinolytic activity were evaluated continuously for 7days using CT scans and CSF D-dimer levels. The primary outcome was 30day mortality. Overall 30day mortality was 26% (n=11), with eight deaths (72.7%) after 3days (D3). Graeb score and CSF D-dimer on admission (D0) were not significantly different between survivors and non-survivors. The temporal profiles of both parameters were distinctly different, with a downward trend in survivors and an upward trend in non-survivors. A mortality rate of 54% was observed between D0-D3 when both scores increased during this interval. In contrast, the mortality was only 4% when both measures decreased during this interval. Early phase (D0-D3) CSF D-dimer or Graeb score change demonstrated high sensitivity of 88% and specificity of 81% when predicting 30day mortality. Early phase CSF D-dimer change in patients with both ICH and IVH is accurate in predicting mortality and may be utilized as a cost-effective surrogate indicator of IVH severity. Serial monitoring of CSF D-dimer dynamic changes is useful for early identification of patients with hematoma progression and poor outcome.