Abstract
Background
Intravenous fluid (IVF) is a frequently recommended intervention in Ebola virus disease (EVD), yet its impact on patient outcomes remains unclear.
Methods
This retrospective cohort study evaluated patients with EVD admitted to 5 Ebola treatment units (ETUs) in West Africa. The primary outcome was the difference in 28-day survival between cases treated and not treated with IVF. To control for demographic and clinical factors related to both IVF exposure and survival, cases were compared using propensity score matching. To control for time-varying patient and treatment factors over the course of ETU care, a marginal structural proportional hazards model (MSPHM) with inverse probability weighting was used to assess for 28-day survival differences.
Results
Among 424 EVD-positive cases with data for analysis, 354 (83.5%) were treated with IVF at some point during their ETU admission. Overall, 146 (41.3%) cases treated with IVF survived, whereas 31 (44.9%) cases not treated with any IVF survived (P = .583). Matched propensity score analysis found no significant difference in 28-day survival between cases treated and not treated with IVF during their first 24 and 48 hours of care. Adjusted MSPHM survival analyses also found no significant difference in 28-day survival for cases treated with IVF (27.3%) compared to those not treated with IVF (26.9%) during their entire ETU admission (P = .893).
Conclusions
After adjustment for patient- and treatment-specific time-varying factors, there was no significant difference in survival among patients with EVD treated with IVF as compared to those not treated with IVF.
Keywords: Ebola virus disease, survival, intravenous fluid, West Africa, marginal structural models
After adjustment for patient- and treatment-specific time varying factors, there was no significant difference in 28-day survival among patients with Ebola virus disease treated with intravenous fluid compared with those not treated with intravenous fluid in a low-resource setting.
(See the Editorial Commentary by Jacob and Brown on pages 1048–9.)
The 2013–2016 Ebola virus disease (EVD) epidemic in West Africa remains the largest recorded since the disease was discovered [1]. The magnitude and high mortality of this outbreak, coupled with the recent 2018 outbreaks in the Democratic Republic of the Congo (DRC), reinforce the need for research on EVD management [2]. As there are no approved treatments for EVD, supportive care constitutes the foundation of clinical management and is an important research focus [3–8].
High-resource settings report lower mortality among patients with EVD compared to low-resource settings, suggesting that high-quality supportive care may improve survival [3, 9–13]. Fluid replacement therapies in particular have been recommended in EVD to combat hypovolemia and organ dysfunction associated with gastrointestinal losses and sepsis [14, 15]. EVD treatment guidelines recommend both oral rehydration solution (ORS) and intravenous fluid (IVF) based on clinical indications [6–8]. However, the evidence evaluating the impact of IVF on outcomes in EVD patients remains limited and conflicting [5, 9, 16–24]. Given the resource limitations present in EVD outbreaks, elucidating the causal effect of IVF on patient outcomes in EVD is of utmost importance [25, 26].
During the West Africa outbreak, the International Medical Corps (IMC), in cooperation with local and global partners, operated 5 Ebola treatment units (ETUs) in Sierra Leone and Liberia. More than 2750 patients presented to these ETUs during the epidemic, of which 20% were diagnosed with EVD [25]. Standardized guidelines (Supplementary Appendix 2) were used to guide management across the ETUs [6, 7]. However, the actual care delivered varied based on individual patient and provider factors and the local availability of resources, including both human and material resources. This study evaluated the impact of IVF on 28-day survival among patients with EVD admitted to these ETUs by controlling for both baseline and time-varying clinical and treatment characteristics over the course of a patient’s ETU care.
METHODS
Study Design, Setting, and Population
This retrospective cohort study utilized data collected as part of routine care at 5 ETUs managed by IMC during their period of operation between 15 September 2014 and 31 December 2015 in Liberia and Sierra Leone [25]. The Sierra Leone Ethics and Scientific Review Committee and the University of Liberia and Rhode Island Hospital institutional review boards provided ethical approval for this study.
All patients admitted to the 5 study ETUs with a final diagnosis of EVD were eligible for inclusion. Patients who were dead on arrival, confirmed to be EVD negative by real-time reverse transcription polymerase chain reaction (RT-PCR) testing, or missing documentation on IVF treatment were excluded from analysis.
Data Collection
Trained clinical providers collected data on patient demographics and baseline clinical signs/symptoms at triage on standardized forms, as previously described [25]. Clinical characteristics and treatment data were recorded 1–6 times daily on all admitted patients based on human resource availability using standardized forms, while final disposition and diagnosis were recorded on standardized discharge forms, as described previously [25, 27]. Ebola virus RT-PCR cycle threshold (Ct) values, which are inversely proportional to viral load, were obtained using previously published procedures [27]. All data forms were digitized and filed into a unified relational electronic research database as described previously [25]. Data quality was measured using lot quality assurance sampling, which demonstrated 99% consistency with the primary ETU records [25, 28].
Statistical Analysis
Descriptive Statistics
Descriptive analyses were performed using frequencies with percentages, median with interquartile range (IQR), or mean with standard deviation (SD), as appropriate. To allow for time-varying analyses, clinical characteristics reported at any point during the day of analysis were considered present for the entire day.
The primary outcome of interest was 28-day survival from the date of ETU admission [27]. Transferred patients for whom outcome data were not available were censored at the time of ETU departure. Cases alive and in care at 28 days were censored at that time. As patients were only discharged from each ETU after they had recovered fully and tested negative for EVD, and as the study ETUs were the only facilities in their respective regions, survival times for cases discharged alive before 28 days were set at 28 days for those who did not return for care after discharge [29].
The primary predictor variable was treatment with IVF, coded as a dichotomous variable (given or not given) for each day of care. Univariate analyses compared demographic, clinical, and treatment characteristics based on treatment with IVF using Pearson χ2 or Fisher exact tests for categorical variables and by Mann-Whitney or t test for continuous variables, as appropriate. IVF type, day of initiation, duration of exposure, and mean daily volumes were summarized.
Propensity Score Matching
To control for confounding by indication, standard propensity score models (PSMs) were used to match patients treated and not treated with IVF during both their first 24 and 48 hours of admission. Generalized additive models were trained using a broad selection of covariates (age, Ct value, bleeding, coma, confusion, dyspnea, anorexia, dysphagia, vomiting, diarrhea, ceftriaxone, cefixime, ondansetron, acetaminophen [paracetamol], artemether, omeprazole, metoclopramide, vitamin C, vitamin A, and multivitamins) that were chosen based on their univariate relationship to IVF exposure and survival in either the study cohort or prior literature [30]. The factors were modeled as binary except for age, where a cubic spline was used to control for the known quadratic relationship between age and survival in EVD [27]. IVF-treated and -untreated cases were then matched based on the nearest propensity score within a caliper of one-third of an SD of the propensity score and exactly matched on Ct value due to its high correlation with the outcome of interest [10, 27, 31]. Covariate balance between treatment and control cohorts was assessed before and after matching. Evaluated on the standardized bias, a threshold of <0.25 was used to ensure balanced treatment groups. After matching, the probability of survival between cases treated and not treated with IVF was assessed using a McNemar χ2 test.
Marginal Structural Proportional Hazards Model
Marginal structural proportional hazards models (MSPHMs) with inverse probability weighting were further employed to control for time-varying patient characteristics and treatments [32–34]. Weights were developed for each patient for each day of follow-up based on the covariates listed above that were inversely proportional to their likelihood of receiving IVF on that day and were used in the score function in the proportional hazard model to reweight both individuals at risk for death and individuals who died on that day. To control for secular outbreak trends that may have influenced the relationship between IVF initiation and survival, the MSPHM was also fit with an additional covariate based on time from opening of the first ETU. Using these weights, a MSPHM for survival was fit and used to estimate the survival functions between 0 and 28 days for cases treated and not treated with IVF based on the day of IVF initiation and duration of treatment. Subsequently, adjusted overall 28-day survival based on the MSPHM was calculated and compared for IVF-treated and -untreated cases using bootstrapping to generate P values to evaluate for significant risk differences between the treatment groups [35]. Unadjusted and adjusted daily survival proportions were plotted between cases treated and not treated with IVF.
Missing Data
The primary missing data were Ct values, as demographic and clinical data were available for nearly all included patients. Since laboratory data were not missing at random (patients who died were more likely to be missing laboratory results), Ct values were collapsed into 3 categories: >22 (low viral load), ≤22 (high viral load), and missing Ct values, so as not to bias analyses (cut-points were based on those used previously) [36].
Secondary Analyses
To determine whether IVF administration had a differential effect in more severely ill patients, an a priori subgroup analysis based on admission Ct value using the derived MSPHM was also performed. To evaluate the impact of IVF volume on survival, a stratified MSPHM survival analysis was performed, separating patients into 2 categories: ≤20 mL/kg/day and >20 mL/kg/day. Patient weights were estimated from standardized formulae for children and prior setting-specific research for adults, and average daily IVF volumes were calculated for each patient [37, 38]. All statistical analyses were performed using R version 3.3.3 software [39]. Supplementary Appendix 1 contains detailed information on model development and statistical packages used.
RESULTS
Characteristics of the Study Population
Four hundred seventy-eight patients with EVD presented to study ETUs during the time period of operation, of which 424 had complete data for analysis (Figure 1). The mean age was 30.5 (SD, 18.7) years and 59.4% were female. Table 1 provides data on the demographic characteristics of patients and the proportion who developed various clinical signs/symptoms or received specific treatments during ETU admission. Ct values were available for 281 patients, of whom 159 (56.6%) had Ct value ≤22 (high viral load).
Figure 1.
Patient flow diagram. Abbreviations: EVD, Ebola virus disease; IV, intravenous.
Table 1.
Overall Study Cohort Characteristics for Patients With Ebola Virus Disease
| Characteristic | No. (%) |
|---|---|
| Age, y, mean (SD) | 30.5 (18.7) |
| Sex | |
| Female | 252 (59.4) |
| Male | 171 (40.3) |
| ETU location | |
| Bong County, Liberia | 129 (30.4) |
| Margibi County, Liberia | 5 (1.2) |
| Kambia District, Sierra Leone | 33 (7.8) |
| Port Loco District, Sierra Leone | 148 (34.9) |
| Bombali District, Sierra Leone | 109 (25.7) |
| Cycle threshold value | |
| Low (≤22) | 159 (37.5) |
| High (>22) | 122 (28.8) |
| Missing | 143 (33.7) |
| Patients with characteristic during ETU admission | |
| Anorexia | 342 (80.7) |
| Any bleeding | 198 (46.9) |
| Coma | 41 (9.9) |
| Confusion | 56 (13.4) |
| Diarrhea | 362 (85.6) |
| Dysphagia | 248 (58.7) |
| Dyspnea | 204 (48.3) |
| Fever | 325 (76.7) |
| Stomach pain | 326 (76.9) |
| Vomiting | 325 (76.7) |
| Patients receiving treatment during ETU admission | |
| Acetaminophen | 410 (96.7) |
| Artemether | 19 (4.5) |
| Artemether-lumefantrine | 379 (89.6) |
| Artesunate | 24 (5.7) |
| Cefixime | 388 (91.7) |
| Ceftriaxone | 97 (22.9) |
| Ciprofloxacin | 30 (7.3) |
| Metoclopramide | 122 (28.8) |
| Metronidazole | 17 (4.2) |
| Multivitamins | 312 (73.8) |
| Omeprazole | 395 (93.2) |
| Ondansetron | 175 (41.3) |
| Oral rehydration solution | 404 (95.5) |
| Promethazine | 8 (2.1) |
| Vitamin A | 357 (84.2) |
| Vitamin C | 398 (93.9) |
| Zinc sulphate | 55 (13.0) |
Data are presented as no. (%) unless otherwise indicated.
Abbreviations: ETU, Ebola treatment unit; SD, standard deviation.
IVF Treatments and Comparisons
IVF was administered to 354 (83.5%) patients at some point during ETU care, with 348 (98.3%) of these cases receiving lactated Ringer solution. Median duration of IVF treatment was 4 (IQR, 3–6) days, with treated cases receiving a mean of 658 (SD, 300) mL/day. Frequency distributions for the time to initiation and duration of IVF exposure are shown in Figures 2 and 3.
Figure 2.
Days to intravenous fluid (IVF) initiation in patients with Ebola virus disease receiving IVF. Data are only representative of patients who received IVF at some point.
Figure 3.
Total days of intravenous fluid (IVF) treatment among all patients with Ebola virus disease. Data are representative of all patients regardless of treatment decision.
Cases receiving IVF were significantly more likely to have low Ct values, symptoms/signs of anorexia, bleeding, coma, confusion, diarrhea, dysphagia, and vomiting, and to receive antimicrobials, antiemetics, ORS, acetaminophen, and vitamins during their ETU stay (Table 2).
Table 2.
Comparisons of Study Cohort Characteristics by Intravenous Fluid Exposure
| Characteristic | IVF (n = 70) | No IVF (n = 354) | P Value |
|---|---|---|---|
| Age, y, mean (SD) | 31.4 (20.6) | 30.3 (18.3) | .685 |
| Sex | .840 | ||
| Female | 40 (57.1) | 212 (59.9) | |
| Male | 29 (41.4) | 142 (40.1) | |
| ETU location | .314 | ||
| Bong County, Liberia | 26 (37.4) | 103 (29.1) | |
| Margibi County, Liberia | 1 (1.4) | 4 (1.1) | |
| Kambia District, Sierra Leone | 2 (2.9) | 31 (8.8) | |
| Port Loco District, Sierra Leone | 21 (30.0) | 127 (35.9) | |
| Bombali District, Sierra Leone | 20 (28.6) | 89 (25.1) | |
| Cycle threshold value | .027 | ||
| Low (≤22) | 23 (32.9) | 136 (38.4) | |
| High (>22) | 14 (20) | 108 (30.5) | |
| Missing | 33 (47.1) | 110 (31.1) | |
| Patients with clinical characteristics during ETU admission | |||
| Anorexia | 48 (68.6) | 294 (83.1) | .017 |
| Any bleeding | 24 (34.3) | 174 (49.2) | .018 |
| Coma | 1 (1.4) | 41 (11.6) | .000 |
| Confusion | 3 (4.3) | 53 (15) | .007 |
| Diarrhea | 48 (68.6) | 315 (89) | .001 |
| Dysphagia | 31 (44.3) | 218 (61.6) | .009 |
| Dyspnea | 28 (40.0) | 175 (49.4) | .205 |
| Fever | 51 (72.9) | 273 (77.1) | .435 |
| Stomach pain | 48 (68.6) | 277 (78.2) | .100 |
| Vomiting | 41 (58.6) | 283 (79.9) | .001 |
| Patients receiving treatment during ETU admission | |||
| Acetaminophen | 59 (84.3) | 350 (98.9) | .003 |
| Artemether | 0 (0.0) | 18 (5.1) | .049 |
| Artemether-lumefantrine | 58 (82.9) | 322 (91.0) | .094 |
| Artesunate | 0 (0.0) | 23 (6.5) | .010 |
| Cefixime | 56 (80.0) | 332 (93.8) | .013 |
| Ceftriaxone | 0 (0.0) | 95 (26.8) | .000 |
| Ciprofloxacin | 0 (0.0) | 30 (8.5) | .001 |
| Metoclopramide | 7 (10.0) | 115 (32.5) | .000 |
| Metronidazole | 0 (0.0) | 16 (4.5) | .066 |
| Multivitamins | 38 (54.3) | 275 (77.7) | .000 |
| Omeprazole | 56 (80.0) | 338 (95.5) | .005 |
| Ondansetron | 3 (4.3) | 170 (48.0) | .000 |
| Oral rehydration solution | 59 (84.3) | 345 (97.5) | .003 |
| Promethazine | 0 (0.0) | 8 (2.3) | .003 |
| Vitamin A | 49 (70.0) | 307 (86.7) | .004 |
| Vitamin C | 56 (80.0) | 341 (96.3) | .001 |
| Zinc sulphate | 4 (5.7) | 49 (13.8) | .056 |
Data are presented as no. (%) unless otherwise indicated.
Abbreviations: ETU, Ebola treatment unit; IVF, intravenous fluid; SD, standard deviation.
Unadjusted 28-Day Survival
Overall, 177 (41.7%) patients survived to 28 days. Viral load on admission was most strongly correlated with survival: 45.8% of cases with a Ct value >22 survived vs 27.1% with a Ct value ≤22 (P < .001). Survivors were less likely to have developed bleeding, coma, confusion, diarrhea, and dyspnea, but more likely to have received cefixime, omeprazole, ORS, acetaminophen, artemether-lumefantrine, and vitamins during their ETU admission (Table 3). Overall, 31 patients (44.9%) not treated with IVF survived whereas 146 (41.3%) treated with IVF survived (P = .583). For patients who died, mean survival time was 5.7 (95% confidence interval [CI], 5.1–6.7) days among cases not receiving IVF and 5.5 (95% CI, 4.8–6.5) days among cases receiving IVF (P = .257). Figure 4 shows the unadjusted survival through 28 days of follow-up by IVF treatment group.
Table 3.
Comparisons of Study Cohort Characteristics by 28-Day Survival
| Characteristic | Survivors (n = 177) | Nonsurvivors (n = 245) | P Value |
|---|---|---|---|
| Age, y, mean (SD) | 28.7 (15.3) | 31.8 (20.8) | .080 |
| Sex | .488 | ||
| Female | 106 (59.9) | 145 (59.2) | |
| Male | 70 (39.5) | 100 (40.8) | |
| ETU location | .277 | ||
| Bong County, Liberia | 57 (32.2) | 72 (29.4) | |
| Margibi County, Liberia | 4 (2.3) | 1 (0.4) | |
| Kambia District, Sierra Leone | 10 (5.7) | 23 (9.4) | |
| Port Loco District, Sierra Leone | 61 (34.5) | 85 (34.7) | |
| Bombali District, Sierra Leone | 45 (25.4) | 64 (26.1) | |
| Cycle threshold value | .000 | ||
| Low (≤22) | 48 (27.1) | 110 (44.9) | |
| High (>22) | 81 (45.8) | 41 (16.7) | |
| Missing | 48 (27.1) | 94 (38.4) | |
| Patients with clinical characteristics during ETU admission | |||
| Anorexia | 142 (80.2) | 196 (80.0) | .963 |
| Any bleeding | 64 (36.2) | 132 (53.9) | .000 |
| Coma | 1 (0.6) | 40 (16.3) | .000 |
| Confusion | 13 (7.3) | 42 (17.1) | .002 |
| Diarrhea | 140 (79.1) | 220 (89.8) | .003 |
| Dysphagia | 99 (55.9) | 148 (60.4) | .350 |
| Dyspnea | 58 (32.8) | 145 (59.2) | .000 |
| Fever | 134 (75.7) | 187 (76.4) | .822 |
| Stomach pain | 134 (75.7) | 189 (77.1) | .746 |
| Vomiting | 133 (75.1) | 189 (77.1) | .649 |
| Patients receiving treatment during ETU admission | |||
| Acetaminophen | 175 (98.9) | 231 (94.3) | .002 |
| Artemether | 3 (1.7) | 14 (5.7) | .039 |
| Artemether-lumefantrine | 167 (94.4) | 210 (85.7) | .003 |
| Artesunate | 9 (5.1) | 13 (5.3) | .936 |
| Cefixime | 170 (96.0) | 215 (87.8) | .003 |
| Ceftriaxone | 34 (19.2) | 61 (24.9) | .143 |
| Ciprofloxacin | 14 (7.9) | 16 (6.5) | .496 |
| Metoclopramide | 48 (27.1) | 72 (29.4) | .544 |
| Metronidazole | 9 (5.1) | 8 (3.3) | .269 |
| Multivitamins | 139 (78.5) | 171 (69.8) | .038 |
| Omeprazole | 175 (98.9) | 215 (87.8) | .000 |
| Ondansetron | 82 (46.3) | 91 (37.1) | .054 |
| Oral rehydration solution | 177 (100) | 224 (91.4) | .000 |
| Promethazine | 3 (1.7) | 4 (1.6) | .904 |
| Vitamin A | 160 (90.4) | 193 (78.8) | .001 |
| Vitamin C | 173 (97.7) | 222 (90.6) | .002 |
| Zinc sulphate | 18 (10.2) | 36 (14.7) | .195 |
Data are presented as no. (%) unless otherwise indicated.
Abbreviations: ETU, Ebola treatment unit; SD, standard deviation.
Figure 4.
Observed survival in patients with Ebola virus disease stratified by treatment or no treatment with intravenous fluid (IVF).
Propensity Score Model
During the first 24 hours of care, 167 (39.4%) cases received IVF whereas 257 (60.6%) did not. A PSM controlling for covariates present during the first 24 hours of admission resulted in 139 case matches between the IVF and no-IVF groups. For matched cases, 28-day survival was 41.0% in cases not treated with IVF and 40.3% in cases treated with IVF during the first 24 hours of ETU admission (P = 1.00). Similarly, during the first 48 hours of care, 257 (60.6%) cases received IVF and 167 (39.4%) did not. A PSM controlling for covariates present during the first 48 hours of admission resulted in 216 case matches between the IVF and no-IVF groups, with minimal standardized bias across all variables (Supplementary Appendix 1). For matched cases, 28-day survival was 46.8% in cases not treated with IVF and 41.2% in cases treated with IVF during the first 48 hours of ETU admission (P = .219).
Marginal Structural Proportional Hazards Model
In adjusted survival analyses using the MSPHM, which allow for controlled comparisons of IVF initiation to no IVF initiation over the entire course of ETU care, no significant difference in 28-day survival was identified. The MSPHM-estimated 28-day survival was 26.9% (95% CI, 16.5%–38.9%) for cases not treated with IVF and 27.3% (95% CI, 20.1%–34.2%) for cases treated with IVF (P = .893). Figure 5 shows the unadjusted survival through 28 days of follow-up by IVF treatment group.
Figure 5.
Adjusted survival of patients with Ebola virus disease, stratified by treatment or no treatment with intravenous fluid (IVF) (marginal structural proportional hazards model with inverse probability weighting).
Subgroup analysis by initial Ct value also found no significant difference in survival based on IVF treatment in patients with high, low, or missing Ct values. In the stratified analysis, 28-day survival was 35.0% (95% CI, 22.2%–47.9%) in patients receiving low daily IVF volumes (≤20 mL/kg/day) and 7.6% (95% CI, 2.1%–15.4%) in patients receiving high daily IVF volumes (>20 mL/kg/day).
DISCUSSION
This study provides the highest-quality evidence to date on the impact of IVF on survival among patients with EVD in a low-resource, outbreak setting. The analyses demonstrate that, when controlling for baseline and time-varying clinical characteristics and treatments, there were no significant differences in survival outcomes based on IVF treatment overall.
The prior evidence of the impact of IVF on survival in EVD is limited and contradictory. During the 1995 EVD epidemic in DRC, mortality decreased as the epidemic progressed, which corresponded with implementation of supportive care protocols, including the use of IVF [40]. Similarly, in a study of 581 patients with EVD admitted to a single ETU in Sierra Leone, which instituted a package of interventions that included 1000–1500 mL/day of Ringer lactate solution for all patients, they noted a significant decrease in their case fatality ratio over time [9]. Although the observed decreases in mortality in these studies coincided with greater utilization of supportive care, including IVF, such time-series data are vulnerable to confounding due to the secular outbreak trends. The case fatality ratio for patients with EVD decreased consistently over the course of the West Africa Ebola epidemic, starting at >70% early in 2014 and dropping to 40% by the end of 2015, which we control for in our final model [41, 42]. In contrast, a study of 382 patients in Liberia found significantly higher mortality among patients treated with IVF compared to those who were not treated, though this study did not control for clinical characteristics and most adjunctive treatments [24].
Though not specific to EVD, 2 recent randomized controlled trials found increased mortality in patients with sepsis treated with IVF boluses in low-resource settings [43, 44]. Prior research has also shown that septic patients receiving large volume IVF resuscitation during the early period of care have increased mortality risk [45, 46]. This comports with the findings of our stratified analysis, which demonstrated lower survival in patients receiving high daily volumes of IVF. Similar to sepsis, vascular leak syndrome occurs in EVD [23], and the risk of edema and respiratory failure in EVD patients has been documented [15, 47, 48], especially with IVF administration [13]. As respiratory and hemofiltration support to address the iatrogenic complications of IVF are rarely available in low-resource settings, limiting the use of IVF in patients with EVD (especially in large volumes) and closely monitoring these patients afterward, either in person or remotely, would be reasonable [5, 49, 50]. Further research is needed to determine if subgroups of patients with EVD may benefit from IVF; the type, volume, and rate at which IVF should be given; and the potential benefit of additional electrolyte supplementation.
There were limitations to this study. Although this study used the most advanced statistical techniques available to control for all measured confounders between IVF treatment and survival in patients with EVD, it could not control for unmeasured confounders, which only a randomized controlled trial can do. While our analyses controlled for the use of common supportive care measures, including ORS, data were not available on the specific daily volume of ORS consumed by each patient. The lack of most vital sign and laboratory data, particularly relating to metabolic derangements, limited the ability to control for additional characteristics that have been shown to be predictive of survival, as did our lack of data on time since onset of illness [10]. Unfortunately, aside from temperature, vital signs were not measured regularly at all sites over their entire period of operation, and aside from EVD RT-PCR testing, other laboratory tests were not routinely available in the study ETUs. However, given that most clinicians in the study ETUs would not have had access to additional vital sign or laboratory data at the time they made the decision to give IVF, it is unlikely that they represent unmeasured confounders of the effect of IVF on survival in this study. Similarly, while Ct values were missing for about a third of the patients studied, the clinicians caring for those patients did not have had access to those missing Ct values either, so they cannot be unmeasured confounders.
In the 5 ETUs, patient rounds were limited to 1–2 hours, 1–6 times per day (with a mean of 3 times per day), based on staffing limitations and the heat stress of working in full personal protective equipment. As such, IVF was generally given as a bolus, and there was limited ability to monitor patients during the periods when no clinicians were present in the ETU wards. Given the prior literature from sub-Saharan Africa on adult and pediatric patients with sepsis, which demonstrates increased mortality with rapid IVF infusions [43, 44], it is possible that slower IVF infusions or better monitoring may have led to improved outcomes in the patients given IVF. In addition, it is entirely possible, based on our results, that small daily volumes of IVF may improve survival whereas large daily volumes worsen survival. A larger, prospective, randomized controlled trial will be necessary to definitively answer this question. All cases studied were treated at ETUs managed by IMC, which may threaten the external validity; however, as the care provided at the facilities was based on the international guidelines utilized at most other ETUs during the epidemic, study results are likely generalizable to similar settings with limited resource availability [6, 7].
CONCLUSIONS
The West Africa Ebola epidemic as well as more recent outbreaks in DRC highlight the global health dangers posed by EVD and the need for better understanding of which management strategies improve survival [2]. The current results, from a large multisite epidemic population, provide the highest-quality evidence to date that IVF, by itself, does not improve survival in patients with EVD in a low-resource setting, suggesting limited utility for routine use of a therapy requiring significant resources to administer. Future research, including data from randomized controlled trials, is needed to further elucidate the effect of IVF type, volume, and infusion strategies on survival in patients with EVD and to determine which subgroups of patients may benefit.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Author contributions. A. R. A., T. L., and A. C. L. conceived the study and obtained the research funding. A. R. A., T. L., and A. C. L. supervised the conduct of data collection and management. A. R. A., T. L., D. Y., and A. C. L. developed the statistical plan and carried out the analyses. All authors took part in drafting and revising the manuscript. A. R. A. and A. C. L. take responsibility for the manuscript in its entirety. All authors had full access to all study data and had final responsibility for the decision to submit for publication.
Acknowledgments. The authors thank the International Medical Corps (IMC) and the governments of Liberia, Sierra Leone, and Guinea for contributing data for this research. They also thank all the generous institutional, corporate, foundation, and individual donors who placed their confidence and trust in the IMC and made its work during the Ebola epidemic possible. They thank the US Naval Medical Research Center, Public Health England, and the Nigerian/European Union Mobile Laboratory for providing laboratory support to the IMC Ebola treatment units in Liberia and Sierra Leone and making their data available for this research. Finally, they thank all of the IMC clinical and nonclinical staff in Liberia and Sierra Leone, including the data collection officers at each Ebola treatment unit, without whom these data would not be available for analysis.
Disclaimer. The funding sources had no involvement in the design or conduct of the study or the decision to submit for publication. The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the views of the IMC or any governmental bodies or academic organizations.
Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases at the National Institutes of Health (grant number R03AI132801). T. L. and D. Y. were partially supported by an institutional development award (award number U54GM115677) from the National Institute of General Medical Sciences of the National Institutes of Health, which funds Advance Clinical and Translational Research.
Potential conflicts of interest. The authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1. World Health Organization. Ebola virus disease Available at: http://www.who.int/mediacentre/factsheets/fs103/en/. Accessed 1 August 2018.
- 2. World Health Organization. Situation report on the Ebola outbreak in North Kivu (December 12, 2018) Available at: http://apps.who.int/iris/bitstream/handle/10665/276811/SITREP_EVD_DRC_20181212-eng.pdf?ua=1. Accessed 1 February 2019.
- 3. Bah EI, Lamah MC, Fletcher T, et al. Clinical presentation of patients with Ebola virus disease in Conakry, Guinea. N Engl J Med 2015; 372:40–7. [DOI] [PubMed] [Google Scholar]
- 4. Bausch DG, Feldmann H, Geisbert TW, et al. Winnipeg Filovirus Clinical Working Group Outbreaks of filovirus hemorrhagic fever: time to refocus on the patient. J Infect Dis 2007; 196(Suppl 2):S136–41. [DOI] [PubMed] [Google Scholar]
- 5. Fowler RA, Fletcher T, Fischer WA 2nd, et al. Caring for critically ill patients with Ebola virus disease. Perspectives from West Africa. Am J Respir Crit Care Med 2014; 190:733–7. [DOI] [PubMed] [Google Scholar]
- 6. World Health Organization. Clinical management of patients with viral haemorrhagic fever: a pocket guide for front-line health workers: interim emergency guidance for country adaptation. Geneva, Switzerland: WHO, 2016. [Google Scholar]
- 7. Sterk E; Médicins Sans Frontières; Filovirus haemorrhagic fever guideline. Geneva, Switzerland: Médicins Sans Frontières, 2008. [Google Scholar]
- 8. Lamontagne F, Fowler RA, Adhikari NK, et al. Evidence-based guidelines for supportive care of patients with Ebola virus disease. Lancet 2018; 391:700–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Ansumana R, Jacobsen KH, Sahr F, et al. Ebola in Freetown area, Sierra Leone—a case study of 581 patients. N Engl J Med 2015; 372:587–8. [DOI] [PubMed] [Google Scholar]
- 10. Hunt L, Gupta-Wright A, Simms V, et al. Clinical presentation, biochemical, and haematological parameters and their association with outcome in patients with Ebola virus disease: an observational cohort study. Lancet Infect Dis 2015; 15:1292–9. [DOI] [PubMed] [Google Scholar]
- 11. Qin E, Bi J, Zhao M, et al. Clinical features of patients with Ebola virus disease in Sierra Leone. Clin Infect Dis 2015; 61:491–5. [DOI] [PubMed] [Google Scholar]
- 12. Schieffelin JS, Shaffer JG, Goba A, et al. KGH Lassa Fever Program; Viral Hemorrhagic Fever Consortium; WHO Clinical Response Team Clinical illness and outcomes in patients with Ebola in Sierra Leone. N Engl J Med 2014; 371:2092–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Uyeki TM, Mehta AK, Davey RT Jr, et al. Working Group of the US–European Clinical Network on Clinical Management of Ebola Virus Disease Patients in the US and Europe Clinical management of Ebola virus disease in the United States and Europe. N Engl J Med 2016; 374:636–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Bishop BM. Potential and emerging treatment options for Ebola virus disease. Ann Pharmacother 2015; 49:196–206. [DOI] [PubMed] [Google Scholar]
- 15. Leligdowicz A, Fischer WA 2nd, Uyeki TM, et al. Ebola virus disease and critical illness. Crit Care 2016; 20:217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Jeffs B, Roddy P, Weatherill D, et al. The Medecins Sans Frontieres intervention in the Marburg hemorrhagic fever epidemic, Uige, Angola, 2005. I. Lessons learned in the hospital. J Infect Dis 2007; 196(Suppl 2):S154–61. [DOI] [PubMed] [Google Scholar]
- 17. Kreuels B, Wichmann D, Emmerich P, et al. A case of severe Ebola virus infection complicated by gram-negative septicemia. N Engl J Med 2014; 371:2394–401. [DOI] [PubMed] [Google Scholar]
- 18. Liddell AM, Davey RT Jr, Mehta AK, et al. Characteristics and clinical management of a cluster of 3 patients with Ebola virus disease, including the first domestically acquired cases in the United States. Ann Intern Med 2015; 163:81–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Lyon GM, Mehta AK, Varkey JB, et al. Emory Serious Communicable Diseases Unit Clinical care of two patients with Ebola virus disease in the United States. N Engl J Med 2014; 371:2402–9. [DOI] [PubMed] [Google Scholar]
- 20. Mupapa K, Massamba M, Kibadi K, et al. Treatment of Ebola hemorrhagic fever with blood transfusions from convalescent patients. International Scientific and Technical Committee. J Infect Dis 1999; 179(Suppl 1):S18–23. [DOI] [PubMed] [Google Scholar]
- 21. Parra JM, Salmerón OJ, Velasco M. The first case of Ebola virus disease acquired outside Africa. N Engl J Med 2014; 371:2439–40. [DOI] [PubMed] [Google Scholar]
- 22. Petrosillo N, Nicastri E, Lanini S, et al. INMI EBOV Team Ebola virus disease complicated with viral interstitial pneumonia: a case report. BMC Infect Dis 2015; 15:432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Wolf T, Kann G, Becker S, et al. Severe Ebola virus disease with vascular leakage and multiorgan failure: treatment of a patient in intensive care. Lancet 2015; 385:1428–35. [DOI] [PubMed] [Google Scholar]
- 24. Gignoux E, Azman AS, de Smet M, et al. Effect of artesunate-amodiaquine on mortality related to Ebola virus disease. N Engl J Med 2016; 374:23–32. [DOI] [PubMed] [Google Scholar]
- 25. Roshania R, MM, Dunbar N, Mansary D, et al. Successful implementation of a multi-country Ebola virus disease clinical surveillance and data collection system in West Africa: findings and lessons learned. Glob Health Sci Pract 2016; 4:394–409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Ebola haemorrhagic fever in Zaire, 1976. Bull World Health Organ 1978; 56:271–93. [PMC free article] [PubMed] [Google Scholar]
- 27. Skrable K, Roshania R, Mallow M, Wolfman V, Siakor M, Levine AC. The natural history of acute Ebola virus disease among patients managed in five Ebola treatment units in West Africa: a retrospective cohort study. PLoS Negl Trop Dis 2017; 11:e0005700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Biedron C, Pagano M, Hedt BL, et al. An assessment of lot quality assurance sampling to evaluate malaria outcome indicators: extending malaria indicator surveys. Int J Epidemiol 2010; 39:72–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Waxman M, Aluisio AR, Rege S, Levine AC. Characteristics and survival of patients with Ebola virus infection, malaria, or both in Sierra Leone: a retrospective cohort study. Lancet Infect Dis 2017; 17:654–60. [DOI] [PubMed] [Google Scholar]
- 30. Stuart EA. Matching methods for causal inference: a review and a look forward. Stat Sci 2010; 25:1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Cournac JM, Karkowski L, Bordes J, et al. Rhabdomyolysis in Ebola virus disease. Results of an observational study in a treatment center in Guinea. Clin Infect Dis 2016; 62:19–23. [DOI] [PubMed] [Google Scholar]
- 32. Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology 2000; 11:550–60. [DOI] [PubMed] [Google Scholar]
- 33. Hernán MA, Brumback B, Robins JM. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology 2000; 11:561–70. [DOI] [PubMed] [Google Scholar]
- 34. Hu L, Hogan JW, Mwangi AW, Siika A. Modeling the causal effect of treatment initiation time on survival: application to HIV/TB co-infection. Biometrics 2018; 74:703–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Efron B. Computers and the theory of statistics: thinking the unthinkable. SIAM Review 1979; 21:460–80. [Google Scholar]
- 36. PREVAIL II Writing Group; Multi-National PREVAIL II Study Team; Davey RT Jr, Dodd L, Proschan MA, et al. A randomized, controlled trial of ZMapp for Ebola virus infection. N Engl J Med 2016; 375:1448–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Koné J, TM, Diamouténé R, Traoré E, et al. A local validation of the APLS pediatric age-based weight estimation formula. J Anesth Crit Care Open Access 2017; 8:00320. [Google Scholar]
- 38. Walpole SC, Prieto-Merino D, Edwards P, Cleland J, Stevens G, Roberts I. The weight of nations: an estimation of adult human biomass. BMC Public Health 2012; 12:439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. R Foundation for Statistical Computing Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2016. [Google Scholar]
- 40. Guimard Y, Bwaka MA, Colebunders R, et al. Organization of patient care during the Ebola hemorrhagic fever epidemic in Kikwit, Democratic Republic of the Congo, 1995. J Infect Dis 1999; 179(Suppl 1):S268–73. [DOI] [PubMed] [Google Scholar]
- 41. World Health Organization Ebola Response Team. After Ebola in West Africa—unpredictable risks, preventable epidemics. N Engl J Med 2016; 375:587–96. [DOI] [PubMed] [Google Scholar]
- 42. Garske T, Cori A, Ariyarajah A, et al. Heterogeneities in the case fatality ratio in the West African Ebola outbreak 2013–2016. Phil Trans R Soc B 2017; 372:20160308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Maitland K, Kiguli S, Opoka RO, et al. FEAST Trial Group Mortality after fluid bolus in African children with severe infection. N Engl J Med 2011; 364:2483–95. [DOI] [PubMed] [Google Scholar]
- 44. Andrews B, Semler MW, Muchemwa L, et al. Effect of an early resuscitation protocol on in-hospital mortality among adults with sepsis and hypotension: a randomized clinical trial. JAMA 2017; 318:1233–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Sadaka F, Juarez M, Naydenov S, O’Brien J. Fluid resuscitation in septic shock: the effect of increasing fluid balance on mortality. J Intensive Care Med 2014; 29:213–7. [DOI] [PubMed] [Google Scholar]
- 46. Marik PE, Linde-Zwirble WT, Bittner EA, Sahatjian J, Hansell D. Fluid administration in severe sepsis and septic shock, patterns and outcomes: an analysis of a large national database. Intensive Care Med 2017; 43:625–32. [DOI] [PubMed] [Google Scholar]
- 47. Clark DV, Jahrling PB, Lawler JV. Clinical management of filovirus-infected patients. Viruses 2012; 4:1668–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Perner A, Fowler RA, Bellomo R, Roberts I. Ebola care and research protocols. Intensive Care Med 2015; 41:111–4. [DOI] [PubMed] [Google Scholar]
- 49. Roberts I, Perner A. Ebola virus disease: clinical care and patient-centred research. Lancet 2014; 384:2001–2. [DOI] [PubMed] [Google Scholar]
- 50. Steinhubl SR, Feye D, Levine AC, Conkright C, Wegerich SW, Conkright G. Validation of a portable, deployable system for continuous vital sign monitoring using a multiparametric wearable sensor and personalised analytics in an Ebola treatment centre. BMJ Glob Health 2016; 1:e000070. [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.