Pier-Alexandre Tardif

Evidence of data quality in trauma registries: a systematic review

BACKGROUND Trauma registries are clinical databases designed for quality improvement activities and research and have made important contributions to the improvements in trauma care during the last few decades. The effectiveness of trauma registries in improving patient outcomes depends on data quality (DQ). However, our understanding of DQ in trauma registries is limited. The objective of this study was to review evidence of the completeness, accuracy, precision, correctness, consistency, and timeliness of data in trauma registries. METHODS A systematic review using MEDLINE, EMBASE, Web of Science, CINAHL, and The Cochrane Library was performed including studies evaluating trauma registry DQ based on completeness, accuracy, precision, correctness, consistency, or timeliness. We also searched MEDLINE to identify regional, national, and international trauma registries whose data were used 10 times or more in original studies in the last 10 years; administrators of those registries were contacted to obtain their latest DQ report. Two authors abstracted the data independently. RESULTS The search retrieved 7,495 distinct published articles, of which 10 were eligible for inclusion. We also reviewed DQ reports from five provincial and international trauma registries. Evaluation was mostly based on completeness with values between 46.8% (mechanism of injury) and 100% (age and sex). Accuracy was between 81.0% (operating room time) and 99.8% (sex). No evidence of data precision or timeliness was available. Correctness varied from 47.6% (Injury Severity Score [ISS]) to 83.2% (Glasgow Coma Scale [GCS] score) and consistency between variables from 87.5% (International Classification of Disease—9th Rev.—Clinical Modification [ICD-9-CM]/Abbreviated Injury Scale [AIS]) to 99.6% (procedure time). CONCLUSION In the few studies we identified, DQ evaluation in trauma registries was mostly based on completeness. There is a need to develop a standardized and reproducible method to evaluate DQ in trauma registries. Determinants of DQ and the impact of DQ on trauma registry analyses such as benchmarking with quality indicators should also be explored.

Complications following hospital admission for traumatic brain injury: A multicenter cohort study

Purpose: To evaluate the incidence, determinants and impact on outcome of in‐hospital complications in adults with traumatic brain injury (TBI). Materials and methods: We conducted a multicenter cohort study of TBI patients admitted between 2007 and 2012 in an inclusive Canadian trauma system. Risk ratios of complications, odds ratios of mortality and geometric mean ratios of length of stay (LOS) were calculated using generalized linear models with adjustment for prognostic indicators and hospital cluster effects. Results: Of 12,887 patients, 3.2% had at least one neurological complication and 22.6% a non‐neurological complication. Mechanical ventilation, head injury severity, blood transfusion and neurosurgical intervention had the strongest correlation with neurological complications. Mechanical ventilation, the Glasgow Coma Scale, blood transfusion and concomitant injuries had the strongest correlation with non‐neurological complications. Neurological and non‐neurological complications were associated with a 85% and 53% increase in the odds of mortality, and a 60% and two‐fold increases in LOS, respectively. Conclusions: More than 20% of patients with TBI developed a complication. Many of these complications were associated with increased mortality and LOS. Results highlight the importance of prevention strategies adapted to treatment decisions and underline the need to improve knowledge on the underuse and overuse of clinical interventions. HIGHLIGHTS1/33 and 1/5 TBI admissions develop neurological and non‐neurological complications.50% increase in mortality and 100% increase in LOSTreatments explain more variation in complications rates than patient risk factors.Highlights the importance of prevention strategies adapted to treatment decisionsUnderlines the need to improve knowledge on intervention underuse and overuse

Invitations received from potential predatory publishers and fraudulent conferences: a 12-month early-career researcher experience

Purpose of the study This study aims to describe all unsolicited electronic invitations received from potential predatory publishers or fraudulent conferences over a 12-month period following the first publication as a corresponding author of a junior academician. Study design Unsolicited invitations received at an institutional email address and perceived to be sent by predatory publishers or fraudulent conferences were collected. Results A total of 502 invitations were included of which 177 (35.3%) had subject matter relevant to the recipient’s research interests and previous work. Two hundred and thirty-seven were invitations to publish a manuscript. Few disclosed the publication fees (32, 13.5%) but they frequently reported accepting all types of manuscripts (167, 70.5%) or emphasised on a deadline to submit (165, 69.6%). Invitations came from 39 publishers (range 1 to 87 invitations per publisher). Two hundred and ten invitations from a potential fraudulent conference were received. These meetings were held in Europe (97, 46.2%), North America (65, 31.0%), Asia (20.4%) or other continents (5, 2.4%) and came from 18 meeting organisation groups (range 1 to 137 invitations per organisation). Becoming an editorial board member (30), the editor-in-chief (1), a guest editor for journal special issue (6) and write a book chapter (11) were some of the roles offered in the other invitations included while no invitation to review a manuscript was received. Conclusions Young researchers are commonly exposed to predatory publishers and fraudulent conferences following a single publication as a corresponding author. Academic institutions worldwide need to educate and inform young researchers of this emerging problem.

Hospital length of stay following admission for traumatic brain injury in a Canadian integrated trauma system: A retrospective multicenter cohort study

BACKGROUND Traumatic brain injury (TBI) is the leading cause of disability in children and young adults and costs CAD

Prognostic value of neuron-specific enolase (NSE) for prediction of post-concussion symptoms following a mild traumatic brain injury: a systematic review

3 billion annually in Canada. Stakeholders have expressed the urgent need to obtain information on resource use for TBI to improve the quality and efficiency of acute care in this patient population. We aimed to assess the components and determinants of hospital and ICU LOS for TBI admissions. METHODS We performed a retrospective multicenter cohort study on 11,199 adults admitted for TBI between 2007 and 2012 in an inclusive Canadian trauma system. Our primary outcome measure was index hospital LOS (admission to the hospital with the highest designation level). Index LOS was compared to total LOS (all consecutive admissions related to the injury). Expected LOS was calculated by matching TBI admissions to all-diagnosis hospital admissions by age, gender, and year of admission. LOS determinants were identified using multilevel linear regression. RESULTS Geometric mean total LOS was 1day longer than geometric mean index LOS (12.6 versus 11.7 days). Observed index and ICU LOS were respectively 4.2days and 2.5days longer than that expected according to all-diagnosis admissions. The six most important determinants of LOS were discharge destination, severity of concomitant injuries, extracranial complications, GCS, TBI severity, and mechanical ventilation, accounting for 80% of explained variation. CONCLUSIONS Results of this multicenter retrospective cohort study suggest that hospital and ICU LOS for TBI admissions are 56% and 119% longer than expected according to all-diagnosis admissions, respectively. In addition, hospital LOS is underestimated when only the index visit is considered and is largely influenced by discharge destination and extracranial complications, suggesting that improvements could be achieved with better discharge planning and interventions targeting prevention of in-hospital complications. This study highlights the importance of considering TBI patients as a distinct population when allocating resources or planning quality improvement interventions.

Impact of wearing a helmet on the risk of hospitalization and intracranial haemorrhage after a sports injury

ABSTRACT Background: This systematic review aimed to determine the prognostic value of neuron-specific enolase (NSE) to predict post-concussion symptoms following mild traumatic brain injury (TBI). Methods: Seven databases were searched for studies evaluating the association between NSE levels and post-concussion symptoms assessed ≥ 3 months (persistent) or ≥ 7 days < 3 months (early) after mild TBI. Two researchers independently screened studies for inclusion, extracted data and appraised quality using the Quality in Prognostic Studies (QUIPS) tool. Results: The search strategy yielded a total of 23,298 citations from which 8 cohorts presented in 10 studies were included. Studies included between 45 and 141 patients (total 608 patients). The outcomes most frequently assessed were post-concussion syndrome (PCS, 12 assessments) and neuropsychological performance deficits (10 assessments). No association was found between an elevated NSE serum level and PCS. Only one study reported a statistically significant association between a higher NSE serum level and alteration of at least three cognitive domains at 2 weeks but this association was no longer significant at 6 weeks. Overall, risk of bias of the included studies was considered moderate. Conclusions: Early NSE serum level is not a strong independent predictor of post-concussion symptoms following mild TBI.

Predatory publishers and fraudulent conferences: Perspectives and implications for novice researchers

ABSTRACT Background: Despite their reported protective effect against the occurrence of head injuries, helmets are still used inconsistently in sports in which they are optional. We aimed to assess the impact of helmet use on the risk of hospitalization and intracranial haemorrhage for trauma occurring during sport activities. Methods: Retrospective cohort of all patients who presented themselves, over an 18-month period, at the emergency department of a tertiary trauma centre for an injury sustained in a sport or leisure activity where the use of a helmet is optional. Impact of helmet use was assessed using multivariable regression analyses (relative risks, RR). Results: Among the 1,022 patients included in the study, half were cyclists and 40% were skiers or snowboarders. A total of 40 % of patients wore a helmet at the time of injury, 18% had a head injury, 16% were hospitalized and 13% of patients with a head injury had an intracranial haemorrhage. Among all patients, no association was observed between hospital admission and helmet use. However, helmet use in patients with a head injury was associated with significant reductions in the risks of hospitalization (RR 0.41 [95% CI: 0.22–0.76]) and intracranial haemorrhage (RR 0.28 [95% CI: 0.11–0.71]). Conclusions: Results suggest that, in recreational athletes who sustain a head injury, helmet use is associated with a reduced risk of hospitalization (all sports) and intracranial haemorrhage (cyclists).

LO93: Prognostic value of S-100B protein for prediction of post-concussion symptoms following a mild traumatic brain injury: systematic review and meta-analysis

Shortly after publishing our first article, we (EM, PAT) started receiving daily electronic invitations to submit additional manuscripts to unfamiliar journals and present at questionable conferences. Unfortunately, our experience is not unique. Considering the intense pressure to publish during medical training, limited knowledge of predatory publishing entities, and lack of local institutional policies to guide trainees’ responses to these flattering invitations, this trend is especially concerning. Publishing in medical journals and presenting at health conferences are valued accomplishments across the continuum of medical education. For example, the number of publications is often used as a metric of productivity and postgraduate trainees are expected to possess lengthy bibliographies to compete for positions, promotions, and grants. Specific to novice researchers, their number of research accomplishments is associated with successful applications in competitive residency programs [1]. However, despite the numerous barriers to publishing in well-established journals

Characteristics of patients included and enrolled in studies on the prognostic value of serum biomarkers for prediction of postconcussion symptoms following a mild traumatic brain injury: a systematic review

This systematic review and meta-analysis aimed to determine the prognostic value of S-100β protein to identify patients with post-concussion symptoms after a mild traumatic brain injury (mTBI). A search strategy was submitted to seven databases from their inception to October 2016. Individual patient data were requested. Cohort studies evaluating the association between S-100β protein level and post-concussion symptoms assessed at least seven days after the mTBI were considered. Outcomes were dichotomized as persistent (≥3 months) or early (≥7 days <3 months). Our search strategy yielded 23,298 citations of which 29 studies including between seven and 223 patients (n = 2505) were included. Post-concussion syndrome (PCS) (16 studies) and neuropsychological symptoms (9 studies) were the most frequently assessed outcomes. The odds of having persistent PCS (odds ratio [OR] 0.62, 95% confidence interval [CI]: 0.34-1.12, p = 0.11, I2 0% [n = five studies]) in patients with an elevated S-100β protein serum level were not significantly different from those of patients with normal values while the odds of having early PCS (OR 1.67, 95% CI: 0.98-2.85, p = 0.06, I2 38% [n = five studies]) were close to statistical significance. Similarly, having an elevated S-100β protein serum level was not associated with the odds of returning to work at six months (OR 2.31, 95% CI: 0.50-10.64, p = 0.28, I2 22% [n = two studies]). Overall risk of bias was considered moderate. Results suggest that the prognostic biomarker S-100β protein has a low clinical value to identify patients at risk of persistent post-concussion symptoms. Variability in injury to S-100ß protein sample time, mTBI populations, and outcomes assessed could potentially explain the lack of association and needs further evaluation.

Low-value clinical practices in injury care: a scoping review protocol

Objective Mild traumatic brain injury (mTBI) has been insufficiently researched, and its definition remains elusive. Investigators are confronted by heterogeneity in patients, mechanism of injury and outcomes. Findings are thus often limited in generalisability and clinical application. Serum protein biomarkers are increasingly assessed to enhance prognostication of outcomes, but their translation into clinical practice has yet to be achieved. A systematic review was performed to describe the adult populations included and enrolled in studies that evaluated the prognostic value of protein biomarkers to predict postconcussion symptoms following an mTBI. Data sources Searches of MEDLINE, Embase, CENTRAL, CINAHL, Web of Science, PsycBITE and PsycINFO up to October 2016. Data selection and extraction Two reviewers independently screened for potentially eligible studies, extracted data and assessed the overall quality of evidence by outcome using the Grading of Recommendations Assessment, Development and Evaluation approach. Results A total of 23 298 citations were obtained from which 166 manuscripts were reviewed. Thirty-six cohort studies (2812 patients) having enrolled between 7 and 311 patients (median 89) fulfilled our inclusion criteria. Most studies excluded patients based on advanced age (n=10 (28%)), neurological disorders (n=20 (56%)), psychiatric disorders (n=17 (47%)), substance abuse disorders (n=13 (36%)) or previous traumatic brain injury (n=10 (28%)). Twenty-one studies (58%) used at least two of these exclusion criteria. The pooled mean age of included patients was 39.3 (SD 4.6) years old (34 studies). The criteria used to define a mTBI were inconsistent. The most frequently reported outcome was postconcussion syndrome using the Rivermead Post-Concussion Symptoms Questionnaire (n=18 (50%)) with follow-ups ranging from 7 days to 5 years after the mTBI. Conclusions Most studies have recruited samples that are not representative and generalisable to the mTBI population. These exclusion criteria limit the potential use and translation of promising serum protein biomarkers to predict postconcussion symptoms.