Traumatic Brain Injury in Latin America: Lifespan Analysis Randomized Control Trial Protocol

Abstract

Background

Although in the developed world the intracranial pressure (ICP) monitor is considered “standard of care” for patients with severe traumatic brain injury (TBI), its usefulness to direct treatment decisions has never been tested rigorously.

Objective

The primary focus is to conduct a high quality randomized, controlled trial to determine if ICP monitoring used to direct TBI treatment improves patient outcomes. By providing education, equipment, and structure, the project will enhance the research capacity of the collaborating investigators and will foster the collaborations established during earlier studies (add refs to papers from earlier studies).

Methods

Study centers were selected that routinely treated ICP based on clinical examination and CT imaging using internal protocols. We randomize patients to either an ICP Monitor Group or an Imaging and Clinical Examination Group. Treatment decisions for the ICP Monitor Group are guided by ICP monitoring, based on established guidelines. Treatment decisions for the Imaging and Clinical Examination Group are made using a single protocol derived from those previously being used at those centers.

Expected Outcomes

There are two study hypotheses: 1) Patients with severe TBI whose acute care treatment is managed using ICP monitors will have improved outcomes and 2) incorporating ICP monitoring into the care of patients with severe TBI will minimize complications and decrease length of ICU stay.

Discussion

This clinical trial tests the effectiveness of a management protocol based on technology considered pivotal to brain trauma treatment in the developed world - the ICP monitor. A randomized controlled trial of ICP monitoring has never been performed - a critical gap in the evidence base that supports the role of ICP monitoring in TBI care. As such, the results of this RCT will have global implications regardless of the level of development of the trauma system.

Rationale and Background Information

ICP monitoring per se has never been subjected to a prospective randomized clinical trial (RCT) to establish its efficacy in improving outcome from sTBI. Hence, there are insufficient data to support its use as a standard of care¹. However, there is a large body of published clinical experience that suggests that ICP monitoring (1) helps in determining prognosis, (2) aids in earlier detection of intracranial mass lesions, (3) is integral to monitoring and management of CPP, (4) may limit the indiscriminate use of therapies to control ICP, which themselves can be potentially harmful, and (5) may improve outcome. ICP monitoring is therefore supported by most head injury experts in the United States and is accepted as a relatively low-risk, high-yield, modest cost intervention. Current evidence-based recommendations¹ are that comatose head injury patients (GCS 3-8) with abnormal CT scans should undergo ICP monitoring (Level II evidence). Comatose patients with normal CT scans have a much lower incidence of intracranial hypertension unless they have two or more of the following features at admission: age over 40, unilateral or bilateral motor posturing, or a systolic blood pressure of less than 90 mm Hg (Level III). ICP monitoring in patients with a normal CT scan with two or more of these risk factors is suggested as an option. Routine ICP monitoring is not indicated in patients with mild or moderate head injury but may be undertaken in certain conscious patients with traumatic mass lesions at the discretion of the treating physician.

Since 2007, the University of Washington (UW), with collaborators from the Oregon Health and Science University, the University of California, San Diego, and Argentina, Bolivia, Ecuador, Colombia, and Brazil, has been conducting an RCT evaluating management of severe TBI using ICP monitoring in the intensive care unit (ICU) at six sites in two countries (ClinicalTrials.gov Identifier: NCT01068522).

In addition, an observational study in five countries has focused on the natural course of recovery and outcome. UW Professor Randall Chesnut, MD is Principal Investigator for the overall research program. Both the study and this protocol are funded by an NIH NINDS/Fogarty sponsored grant “Traumatic Brain Injury in Latin America: Lifespan Analysis”. The studies are run in collaboration with Fundacion ALAS (Apoyo al Lesionado Neurologico Agudo – “Support to the Acute Neurological Patient”), and the Latin American Brain Injury Consortium (LABIC). As part of the study implementation strategy, we established a local data center in Rosario (CIIC - Centro de Infomatica y Investigaciones Clinica), which performed site training and oversight as well as local data management. Only the RCT protocol is described here.

Study Goals and Objectives

The primary focus for this scientific investigation is to conduct a high quality RCT to determine if ICP monitoring to direct treatment of patients with TBI improves medical practice and patient outcomes.

Specific Aim: In a randomized controlled trial in trauma centers in Bolivia and Ecuador, test the effect on outcomes of management of severe TBI guided by information from ICP monitors vs. a standard protocol based on imaging and clinical examination.

Hypothesis #1: Patients with severe TBI whose acute care treatment is managed using ICP monitors will have significantly lower mortality and better neuropsychological and functional recovery over 6 months post-trauma than those whose treatment is managed with the standard protocol.

Hypothesis #2: The incorporation of ICP monitoring into the care of patients with severe TBI will minimize complications and decrease length of stay in ICU.

Study Design

Protocol

This study is a randomized, outcome-masked, clinical trial of the management of sTBI, with or without monitoring of intracranial pressure. The study is designed to test the effect on outcomes of management in 324 participants at centers in Bolivia and Ecuador and is a 2-group, parallel design. Within 24 hours of injury (or 24 hours of deterioration, but no later than 72 hours after injury), patients with severe TBI are randomized to brain injury management based on ICP monitoring (“ICP group”) or brain injury management conducted according to a protocol that does not include monitoring of ICP, (“Imaging and Clinical Examination group – ICE group”). Patients are followed and their outcome evaluated by a blinded examiner at 3 and 6 months after injury.

Methodology

Study Sites

The study was initiated at three centers in Bolivia. Two were located in Santa Cruz de la Sierra, Boliva (Hospital San Juan de Dios and Hospital Japones) and one in Cochabamba, Bolivia (Hospital Viedma). As the study progressed, we incorporated three additional centers: one in Tarija, Bolivia (Hospital San Juan de Dios), one in Quito, Ecuador (Hospital Eugenio Espejo), and one in Guayaquil, Ecuador (Hospital Luis Vernaza).

Study Enrollment Procedures

Identifying and recruiting candidates for the trial

When a TBI patient is received in the participating hospital, he/she is evaluated by an emergency department physician to determine whether they fulfill the eligibility criteria listed below. Immediate contact is made with the 24-hour-on-call study coordinator, who confirms eligibility and requests consent from the patient’s legally authorized representative.

Eligibility Criteria

Inclusion Criteria

• Admission to study hospital within 24 hours of injury

• Closed head trauma

• Glasgow Coma Scale (GCS) ≤ 8 or if intubated, GCS Motor ≤ 5 on admission, or deterioration to that level within the first 48 hours after injury

• No foreign object in brain parenchyma

• Randomized:

  • Within 24 hours of injury [for patients with GCS ≤ 8 on admission] or

  • Within 24 hours of deterioration [patients deteriorating to GCS ≤ 8 within 48 hours of injury]

• Age > 12

Exclusion Criteria

• GCS of 3 with bilateral fixed and dilated pupils and/or decision to not actively treat prior to enrollment into study

• No beds available in ICU

• No ICP monitor available

• Pregnancy

• Prisoner

• No consent

• Non-survivable injury

• Other (e.g., Pre-injury life expectancy under 1 year)

• Pre-existing neurological disability that would confound outcome

Consent

Informed consent is obtained by the study coordinator with the site PI. As the patients are in coma, consent for those > 20 years of age is discussed with the family; for those 13 to 20 years, discussion is with the parent or legal guardian. Written and verbal consent are obtained in Spanish unless the consentees spoke only Aymaran or Quechuan, which have no common written form, wherein consent is done verbally. Care was taken to include in this discussion not only study issues but also the concept of clinical research, which is quite uncommon in these areas. Whenever the patient regains consciousness, the consent process is redone to obtain consent (or Child Assent for those 13 to 20 years) to continue.

Randomization

The randomization lists, computer-generated by the statistician/database manager at the University of Washington Data Center (UWDC), are blocked for balance and stratified by site, age (≤40 vs. >40), and GCS (3–5 vs 6–8, or GCS motor 1–2 vs GCS motor 3–5 if intubated). Randomization is routinely implemented using a password-protected application on a laptop at each site. If the computer is not available when it is time to randomize a patient, the study coordinator calls a study monitor at Fundacion ALAS who flips a coin and tells the site the group assignment. The Monitor sends an e-mail indicating the assignment to the UWDC as well as the site to confirm the assignment.

Study Interventions

Interventions, Administration, and Duration

Treatment arms

There are two arms in this study, the ICP Monitor Group and the Imaging and Clinical Examination Group. Management of patients who are randomized to the ICP Monitor Group was based specifically on the presence of intracranial hypertension and followed the Guidelines for the Management of Severe Brain Injury¹.

Since there is no literature base from which to derive a management protocol for sTBI patients being treated without ICP monitoring, we developed an ad hoc protocol in conjunction with the investigators at the three initial sites. Fortunately, they had developed similar treatment approaches at their institutions and it was possible to formulate a protocol acceptable to all. One criterion for subsequent acceptance of institutions as study centers was acceptance of this protocol by all involved in the care of sTBI patients. (In the absence of a supporting literature base or common consensus in this area, this acceptance criterion was not trivial and resulted in exclusion of one group).

Overall Treatment Guidelines (both groups)

During this study, we strongly encouraged the following treatment approaches whenever available and/or possible:

1. Patient monitoring measures

   1. Place patient on mechanical ventilation (VM)

   2. Place continuous SaPO2 monitor and EtCO2 monitors

   3. Insert indwelling urinary catheter to monitor urine output

   4. Insert arterial catheter for arterial mean pressure monitoring

   5. Insert central venous catheter for infusion of solutions and central venous pressure monitoring.

   6. Monitor neurological clinical status each hour

      • Pupils

      • GCS

      • Vital signs, etc.

   7. Brain CT (see below)

2. General measures

   1. Head positioning 30°

   2. Head and neck in neutral position and aligned

   3. Avoid hyperthermia

      1. Defined as central temperature ≥ 38 °C

         1. Non-drug measures (cooling)

         2. Dipirona (Metamizole sodium)

   4. Early enteral nutritional support

      1. Before 48 hours

      2. 25 Kcal/kg weight

   5. Pharmacologic prophylaxis of post-traumatic seizures

      1. Phenytoin (IV or PO)

         1. Load and maintenance dose as is being given in each hospital

   6. Gastric bleeding prophylaxis

      1. Ranitidine or Omeprazol

   7. Avoid decubitus lesions

   8. Deep venous thrombosis prophylaxis

   9. Frequent tracheal suctioning with sterile technique to prevent pulmonary infections

   10. Maintain Hb ≥ 7 mg/dL

3. Routine CT scans

   1. First CT: on Hospital admission

   2. Second CT: 48 hours after the first CT

   3. Third CT: 5–7 days after the first CT

Imaging and Clinical Examination Group

Treatment decisions for the Imaging and Clinical Examination Group are made following the protocol developed for this study (Figures 1 and 2). A detailed description of the protocol manual of procedures is found at the study website: www.globalneurotrauma.org.

Figure 1.

Schematic for Imaging and Clinical Examination Group treatment. The neuroworsening protocol is in Figure 2.

Figure 2.

Schematic for the euroworsening protocol for the Imaging and Clinical Examination Group.

ICP Monitor Group

Treatment decisions for the ICP Monitoring group are made following protocols based on the Guidelines for the Management of Severe Traumatic Brain Injury¹. A tiered approach to interventions was followed, with lower risk therapies used before riskier ones. A detailed description of the protocol MOP is found at the study website: www.globalneurotrauma.org.

Outcomes

Primary outcome

The primary outcome is a composite measure based on individual outcome variables indicated below. For each measure, the participants are ranked from 1 (worst) to n (best). Those who died are given the lowest ranking and, for the neuropsychological measures, those who are so neurologically impaired that they cannot perform the test are given the next lowest rankings. Tied ranks are all given the average of the tied ranks. For example if there are 3 deaths, all are given the rank 2 (the average of 1, 2, and 3). To increase interpretability, the ranks are converted to percentiles, giving the percent at or worse than the participant’s score. For each person, the percentiles on the different measures are averaged.² Outcome Variables (see Table 1)

Table.

Measures of patient outcome

General	3 months post injury	6 months post injury
Mortality	GOAT	GOAT
Time to follow commands	GOS-E2	GOS-E
GOAT1 at discharge	DRS3	DRS
Length of ICU stay5		Neuropsychological Battery
Complications5

¹Galveston Orientation Amnesia Test

²Glasgow Outcome Scale - Extended

³Disability Rating Scale,

⁵Outcomes used to test hypothesis 2.

• Mortality

• Time to follow commands (TFC--measured as time from injury to following simple commands as defined by a score of 6 on the motor scale of the GCS)

• The Galveston Orientation and Amnesia Test (GOAT) Sum of Errors³ at discharge, 3 and 6 months after injury

• Functional status at 3 and 6 months (see below for further definition)

• Neuropsychological Test Battery (see below for individual measures).

Functional Status

A portion of the Disability Rating Scale (DRS)⁴ measuring eye opening, communication ability, and motor response, and the Glasgow Outcome Scale Extended⁵ are used to measure the level of functioning in everyday life. These are the two most commonly used measures of functional outcome in traumatic brain injury. Both have been translated and used extensively in previous research in Latin America by this research group.

Neuropsychological Test Battery

A battery of measures that examine important neuropsychological constructs sensitive to the integrity of brain functions, including traumatic brain injury, was used. Cognitive, performance based measures were included because they are sensitive to the effects of TBI and oftentimes the major cause of disabilities.

At discharge and at 3 and 6 months, the Galveston Orientation and Amnesia Test³ (a measure of orientation and amnesia⁶) was administered.

The neuropsychological domains and the measures used at 6 months to examine them as follows:

• Mental State (Mini-Mental State Examination^7,8)

• Working Memory (Paced Auditory Serial Addition Test (First subtest only))⁹

• Speed of Information Processing (WAIS III Digit Symbol, and Symbol Search subtests^{10, 11}; Color Trails part 1¹²; Trail Making Test Part A¹³)

• Episodic Memory and Learning (Spanish Verbal Learning Test¹⁴, Brief Visuospatial Memory Test Revised^13,15,16)

• Executive Functions (Noun Fluency (animals)¹⁷, Verbal Fluency (actions)¹⁸; Controlled Oral Word Association Test (COWAT)¹⁹, Color Trails part 2^12,13)

• Motor Speed & Dexterity (Grooved Pegboard Test)³

Primary Outcome Composite (see Table 1): Scores used in the Composite measure include TFC at discharge, GOAT, GOS-E, and DRS at 3 and 6 months and the neuropsychological battery at 6 months. The neuropsychological battery includes the MMSE total score, the Spanish Verbal Learning Test total learning score and Long Delay Free Recall, the Brief Visuospatial Memory Test Revised total learning number correct and delay correct, WAIS III Digit Symbol and Symbol Search scores, Color Trails 2 time to completion, number correct on PASAT and three subcomposite scores where tests are grouped together to each form 1 variable to be entered into the composite. The first is composed of Grooved Pegboard dominant and non-dominant times. The second subcomposite is composed of Color Trails 1 and Trail Making Test Part A times to completion. The third subcomposite is composed of total correct on COWAT, Category Fluency Test for Animals, and Actions. Before ranking and entering the composite (or subcomposite), each neuropsychological test score will be regressed on age, sex and years of education, factors which are known to effect neuropsychological performance, to decrease variability. The residuals are ranked and entered onto the composite or subcomposite. Use of T-scores based on the norms for monolingual Spanish-speakers was considered but discarded because uninjured Bolivians did not have scores with the expected mean of 50 and there was a substantial relationship between years of education and the T-scores for some measures.

Discussion

Globalization encourages international scientific collaborations and the successful implementation of this RCT underscores both the capacity and willingness of LMIC investigators to conduct rigorous Class I trials. Challenges included occasional language/translation barriers, import/export delays, communication challenges, transaction fees associated with multiple currencies, and additional levels of review associated with multinational research. An overview of operational issues relevant to conducting RCTs in LMICs will be published separately.

Trial Status

Follow up visits and data collection were ongoing at the time this manuscript was written.

Safety Considerations

Data and Safety Monitoring Plan

The PIs at each site are notified immediately by the study coordinators of any serious adverse event (SAE) and evaluate whether adverse events are unusual given the patient’s injury. All SAEs are reported to the local IRBs and to the Data Coordinating Center.

Serious Adverse Event Reporting

All serious adverse events (SAEs) data are recorded and analyzed to identify safety problems that may not be readily apparent from case-to-case review of medical information.

DSMB

The study operates under a Data Safety and Monitoring Board that provides overall study oversight and an additional level of patient safety review.

Follow-Up

Subjects were followed for outcome assessments at 3 and 6 months after injury as well as data collection out to 6 months for serious adverse events. Outcome examiners were masked to treatment group. The study provided no additional treatment after the person left the hospital.

Data Management and Statistical Analysis

Data analysis

For hypothesis 1, testing will be by a blocked Wilcoxon test¹⁸ comparing the average percentiles for people in the two treatment groups after controlling for center, TBI severity group and age group. A 2-sided .05 significance level will be used. Analysis will be according to the intention-to-treat principle, i.e., all randomized cases will be followed and included with their assigned treatment group regardless of the management protocol actually used. To supplement the composite test of the overall hypothesis, individual measures will be summarized for each group.

One interim analysis was planned after half the patients reached the 6-month assessment, to determine efficacy. The interim analysis used O’Brien-Fleming²⁰ boundaries to decide whether to terminate because of a positive effect of ICP-guided management. In addition, to terminate early because of efficacy, analysis of at least one of the individual outcomes – mortality or Glasgow Outcome Scale Extended-- must have indicated significant improvement with ICP-guided management at the one-sided nominal 0.05 level. Because this is a trial for which, in the US at least, there is strong feeling that the experimental treatment (ICP monitoring) is superior, the study would not be stopped early for futility, that is because it is unlikely to show a significantly positive effect of ICP monitoring. If that is the case, the study needs as narrow a confidence limit as possible, so stopping early would be contraindicated. These criteria were not met at the interim analysis and the study has been allowed to continue.

For hypothesis 2, length of stay in the ICU will be analyzed using a blocked Wilcoxon test as above. The most frequent complications (major respiratory problems, sepsis, decubitus ulcers) as well as any non-neurologic complication will be summarized as present/absent and analyzed separately using a Mantel-Haenszel test with stratification as above. A 2-sided significance level of .01 will be used for each test to account for the multiple comparisons.

Power

A sample size of 324 patients (162/group) gives the study 80% power to detect an average improvement of 10 percentage points on the percent with good or moderate recovery on the GOS accompanied by similar improvement on the other outcome measures.

Quality Assurance

This team has substantial experience in all elements necessary to successfully conduct high quality randomized control trials – representative samples, adequate random assignment, outcome assessment blinding, high follow-up rate, and attention to potential errors. Latin American key investigators new to the team were formally trained in clinical research, including, for some, Master Degree and year-long certificate programs, and clinical research rotations. Data quality was managed by accuracy checks of every data point and tallying errors until an acceptable level was reached and then maintained through systematic audits. In addition to weekly conference calls among members of the research and administrative team, we hold annual all-team meetings to reinforce training, troubleshoot problems and encourage collective ownership and responsibility. Monthly phone calls are held with outcome examiners to provide continued feedback to encourage high follow-up rates and completeness and accuracy of results. These efforts have produced good results including data error rates below 1% and a current follow-up rate of 92%. In addition to four study-appointed clinical and data quality monitors, external review occurs through regular meetings of the DSMB to monitor the study overall.

Expected Outcomes of the Study

This clinical trial tests the effectiveness of management based on technology considered pivotal to brain trauma treatment in the developed world - the ICP monitor. An RCT of ICP monitoring has never been performed. As noted in every edition of the Guidelines for the Management of Severe Brain Injury^1,21–23 this is a critical gap in the evidence base that supports the role of ICP monitoring in TBI care. As such, the results of this RCT will have global implications regardless of the level of development of the trauma system. It is anticipated that the results will be published during the final quarter of 2012. They are to be presented during the annual meetings of the Congress of Neurological Surgeons and the Neurocritical Care Society in October of 2012.

Duration of project: April, 2007 – March, 2013

Project Management – Roles and Responsibilities

Randall Chesnut, MD, FCCM, FACS, Principal Investigator and Medical Director

Nancy Temkin, PhD, Biostatistician

Nancy Carney, PhD, Project Manager

Sureyya Dikmen, PhD Neuropsychologist

Jim Pridgeon, MHA, Seattle Study Administrator, Editor

Joan Machamer, MA, Outcome measures coordinator

Jason Barber, MS, Biostatistician/Database Manager

Juanita M. Celix, MD. ICP Monitor implementation training

Kelley Chaddock, BA, Research Coordinator

Marianna Cherner, PhD, Neuropsychologist, Outcome measures consultant (Spanish)

Terrence Hendrix, BA, Latin America site outcomes coordinator

Latin American Investigators:

Carlos Rondina, MD, Latin American PI, President Fundacion ALAS

Silvia Lujan, MD, Site personnel trainer/coordinator/monitor

Gustavo Petroni, MD Site personnel trainer/coordinator/monitor

Walter Videtta, MD, Co-Investigator, President Latin American Brain Injury Consortium

Study Sites

Bolivia:

Dr. Victor Severo Alanis Mirones, Site PI, Hospital San Juan De Dios, Santa Cruz, Bolivia

Dr. Gustavo La Fuente Zerain, Site PI, Hospital Japones, Santa Cruz, Bolivia

Dr. Arturo Lavadenz Cuentas Viedma Hospital, Site PI, Cochabamba, Bolivia

Dr. H. Roberto Merida Maldonado, Site PI, San Juan de Dios Hospital, Tarija, Bolivia

Dr. Fredy Sandi Lora, Country coordinator, La Paz, Bolivia

Ecuador:

Dr. Edison Manuel Jibaja Vega, Site PI, Espejo Hospital, Quito, Ecuador.

Dr. Luis González, Site PI, Vernaza Hospital, Guayaquil, Ecuador.

Ethics

For over two decades, calls for an RCT on ICP monitoring were countered by ethical concerns about an unmonitored control group. Having personally expressed such concerns, the authors were apprised of the HIC-centered nature of this objection when, in the process of developing a separate prospective sTBI outcome study, they encountered Latin American Intensivists and Neurosurgeons who rarely used such monitoring and were at equipoise about its utility. Ventricular-catheter-based monitoring was universally available so resource limitations did not play an absolute role in this choice. They did not feel that this practice would alter significantly over the next five years and expressed interest in establishing the efficacy of ICP monitoring in their setting. Reflecting this equipoise, we designed this RCT. It was subsequently reviewed and accepted by the NIH, the University of Washington IRB, and the FWA-approved ethics committees/IRBs at all study centers.