The 2010 Haiti earthquake affected 3 million individuals, killing and injuring uncountable. Most healthcare facilities and many healthcare providers were affected themselves. A quarter of the population was relocated in temporary shelters 1.
This disaster resulted in an immediate global humanitarian response. The Belgian First Aid and Support Team (B-FAST), with trained rescue, medical and logistic professionals cooperating with diplomatic and security personnel, was the first international team with an operational field hospital (FH) in the area 2. From day 2 after the event, the team provided acute care to about 7000 patients during the first 4 weeks after the earthquake, retrieving medical supplies from preconfigured Interagency Emergency Health Kits (IEHK) 3.
A prerequisite to adopting an evidence-based approach in humanitarian response is the need to assemble solid evidence from the results of relevant empirical studies 4. Collection of reliable data will always be difficult in emergency situations as healthcare providers prioritize treating casualties over documenting 5–8. Published reports generally elucidate immediate effects (estimated numbers of dead, injured and displaced) and late consequences such as outbreaks and mental health problems 9,10. Little research has examined the transition from acute response to the recovery phase, merging disaster-related features with pre-existing local pathology and health issues secondary to disruption of the healthcare system, poor sanitary conditions and crowded internally displaced person (IDP) camps 7,10–13. This phase occurs when most disaster-relief teams have left and there is no evidence of when a ‘return to baseline pathology’ can be expected 14. This study is the first to compare postearthquake diagnoses with baseline data of the same affected area and time span.
The aim of this study is to document diagnoses in patients presenting to an FH or to outreach teams in IDP camps; to compare those with baseline medical data of patients from the same area during the same time span before and after the event; and to discuss implications and formulate recommendations for future disaster-relief operations 15.
The hypothesis is that besides earthquake-related trauma, medical problems emerge, questioning an appropriate composition of foreign medical teams (FMT) and resources in the IEHK sent to the disaster site.
A retrospective descriptive cohort analysis was carried out on prospectively obtained medical records, collected by B-FAST between 14th January and 2nd February 2010, and compared with data prospectively collected by Médecins Sans Frontières/Doctors Without Borders (MSF) in the corresponding weeks of the year in 2009, 2011 and 2012 (see Table, Supplemental digital content 1, http://links.lww.com/EJEM/A108 situating the corresponding periods).
The study protocol was approved by the Ethical Committee of the Universitair Ziekenhuis Brussel, Belgium.
B-FAST erected an enhanced level I Medical-Surgical FH with units for triage, emergency care, surgery, recovery and a 20-bed hospitalization unit. The B-FAST team comprised an Urban Search And Rescue section, 12 physicians, 34 nurses, two pharmacists and 20 logisticians and incorporated Haitian nurses, translators and clerks (see Table, Supplemental digital content 2, http://links.lww.com/EJEM/A109 listing the B-FAST team composition). Their Outpatient Assistance Team (OAT) provided local healthcare to seven IDP camps.
Patients presented to the FH or the OAT by their own means or were introduced by outreach teams. Triage performed by a trained nurse filtered patients with minimal complaints who were discharged with symptomatic treatment, but not registered. For all patients receiving care in the FH or IDP camps, two WHO health cards were completed, registering age, sex, all physical complaints and one single primary diagnosis. Names were only written on the form provided to the patient, not on the copy retained for this study. Each patient was additionally questioned, examined and diagnosed by physicians, treated on the spot if possible and vaccinated when necessary. Patients were hospitalized if necessary or treated ambulatorily with explanations in their own language. Materials and drugs were retrieved from preconfigured IEHK.
MSF, present in Haiti since 1991, opened the Martissant Medical Services in 2006, with, among other facilities, an Outpatient Department-Emergency Room and a 20-bed hospital. MSF recruited patients from their Outpatient Department-Emergency Room and OATs going into IDP camps. At least two trained physicians and nurses examined these patients, registering age, complaints, main diagnoses, and treatment or follow-up on MSF-adapted WHO patient forms 16. Lab, X-ray device and trained technicians were available 24/7.
Participants: cohort patient selection
Patients from both B-FAST and MSF originated from the same districts in Port-au-Prince: Delmas, Centre-Ville, Carrefour, Cité-Soleil, Bel-Air and Martissant. Both OATs visited the same IDP camps and accepted all patients presenting, but only registered patients were included in the study.
Diagnoses were made on the basis of complaints and physical examination as the B-FAST team had little access to laboratory or imaging diagnostic capability. One single diagnosis per patient was registered according to a list of 40 possible diagnoses adapted from case descriptions in the WHO Field Manual on Communicable disease control in emergencies and the Sphere Project Handbook 16,17 (see Table, Supplemental digital content 3, http://links.lww.com/EJEM/A110 listing all categories and diagnoses case descriptions in detail). These diagnoses were post-hoc categorized on the basis of adapted ICD-10 codes (Table 1) 18. Three extra categories were subanalysed, regrouping similar diagnoses from different categories: earthquake-related injury was divided into accidental trauma (amputations, fractures, contusions) versus acute skin wounds and lacerations to be comparable with the MSF classification; FH patients were subanalysed in terms of trauma versus nontrauma diagnoses; and infectious diagnoses were regrouped as described in Table 1.
All B-FAST data were anonymized according to the Helsinki Convention and entered into SPSS v22.0 (IBM Corporation, Chicago, Illinois, USA). All MSF consultations were anonymously registered in an Excel database, available from the MSF Central Office in Belgium as weekly reports on the total numbers of patients in two age categories (under and over 5) with primary diagnoses on the basis of the WHO Field Manual on Communicable disease control in emergencies 16,17.
Although both B-FAST and MSF recruited patients from the same area, worked in comparable settings (outpatient clinics combined with OAT in the same IDP camps), during corresponding periods (weeks 2–5 of each year), circumstances were different and forms were filled out by different healthcare providers, possibly introducing sampling bias (see Table, Supplemental digital content 4, http://links.lww.com/EJEM/A111 comparing B-FAST and MSF settings).
Statistical methods/data analysis
Patients with missing or unreadable data were excluded. Descriptive statistics for the discrete outcome variables are presented as frequencies [n (%)] broken down for age (<5 and ≥5 years old) and period (4 weeks). Diagnoses of the victims examined by B-FAST during 4 weeks after the 2010 earthquake were compared with the reference data collected by MSF in the corresponding weeks in 2009, 2011 and 2012, using 99% confidence intervals, according to the Wilson ‘Score’ method 19.
B-FAST examined over 7000 patients during the first month after the earthquake, of whom about 3500 were triage admissions. Clinical notes were available from 2931 patients. Excluding incomplete forms, 2795 (1861 in FH and 934 in IDP camps) were included and analysed in this study.
MSF reported on 6407 patients in the corresponding weeks in 2009, 6033 in 2011 and 7300 in 2012 (see Figure, Supplemental digital content 5, http://links.lww.com/EJEM/A112 composition of the database). MSF data (2010) are lacking because the earthquake-hit facility was inoperable during the corresponding period.
Descriptive B-FAST data
Of all 2795 B-FAST patients, 1627 (58%) were females, median age 24 years (range 0–95): 25 (0–95) among females and 21 (0–87) among males. Over 37% (n=1045) were minors (<18); 371 patients (13%) were less than 5 years old.
Out of 1861 FH patients, 1071 (58%) were females, median age 25 (range 0–90), and 223 (12%) were less than 5 years old. Out of 934 IDP patients, 556 (60%) were females, median age 22 years (range 0–95), and 148 (16%) were less than 5 years old (see Figure, Supplemental digital content 6, http://links.lww.com/EJEM/A113 comparing Haitian and B-FAST populations).
Among all B-FAST patients in the 4 weeks after the earthquake, injury (n=1081 or 39%) with accidental trauma (n=584 or 21%) and acute wounds (n=497 or 18%) accounted for the most common diagnoses, but presented almost all in the fortnight after the earthquake (Table 2). As the load of injured patients regressed after 2 weeks, apart from follow-up injury cases (n=257 or 9%), relatively more medical conditions were encountered: respiratory (n=462/17%), digestive (n=298/11%), genitourinary (n=189/7%), ophthalmological (n=117/4%) and dermatological (n=112/4%) diagnoses, producing a shift in the type of pathology (Figs 1 and 2).
Of all B-FAST patients after the earthquake, 1057 (38%) had features of infection. This was more pronounced in the subgroup less than 5 years old (n= 273 or 74% vs. n=784 or 32% in ≥5 years) and the proportion of infectious diseases started increasing particularly 2 weeks after the event, increasing to more than 53% of all patients.
In children under 5, most diagnoses were of respiratory (n=128/35%), digestive (n=107/29%) or dermatological (n=27/7%) origin; 73 (20%) suffered from injuries, of whom most had acute wounds (n=38/10%), and 34 (9%) had accidental trauma.
In most IDP patients (n=934), irrespective of their age, respiratory (n=296/32%), digestive (n=168/18%) or genitourinary (n=112/12%) diseases were diagnosed; 70 (8%) had ophthalmological diseases and 58 (6%) had dermatological diseases. Two weeks after the event, 90 IDP patients still suffered from untreated earthquake-related injury (10%) and 667 (71%) had features of infections.
Comparison between B-FAST and MSF data
The B-FAST diagnosis categories were post-hoc compared with those registered by MSF in the year before and 2 years after the earthquake for the total patient population, for age categories (under and above 5 years old) and per corresponding week of the year 16–18.
Comparing all ages and the entire period of 4 weeks using 99% confidence intervals (Table 3 and Fig. 3), there was significantly less violence and accidental trauma, less general and follow-up cases in the year of the earthquake than before and longer after the earthquake, but there were more acute wounds, and respiratory, ophthalmological, digestive, neurological, dermatological and psychological diagnoses. There were no changes in genitourinary and circulatory diagnoses (see Figures, Supplemental digital content 7, http://links.lww.com/EJEM/A114 and Supplemental digital content 8, http://links.lww.com/EJEM/A115 comparing weekly diagnoses between B-FAST and MSF).
In Haitian adults and children over 5, the earthquake significantly increased the incidence of accidental trauma and acute wounds, respiratory, ophthalmological, genitourinary, neurological and psychological problems. However, there were significantly fewer violence victims than usual.
In children under 5, the earthquake predominantly increased the occurrence of acute wounds, respiratory, digestive and dermatological problems (see Figures, Supplemental digital content 9, http://links.lww.com/EJEM/A116 and Supplemental digital content 10, http://links.lww.com/EJEM/A117 comparison of proportion of diagnoses in both age categories).
Trauma cases represented more than 90% of all patients in the first 2 weeks after the event, of comparable size as reported by other FMT 11,12. In the following 2 weeks, most patients (75%) presented with medical conditions and more than half of all patients showed features of infection, mostly of respiratory and digestive, but also genitourinary and dermatological origin 12,16. This transition from trauma to predominantly nontrauma cases within 2 weeks is shown in Fig. 2.
A possible explanation could be related to the shift of all major hospitals into trauma centres and the deployment of more sophisticated FH, diverting nontrauma patients to lower level FHs. In children under 5, respiratory and digestive problems emerged even more to the prejudice of trauma-related injuries as children have a lower survival rate after serious injury 1.
Comparison between the B-FAST data and the MSF baseline data from before and after the earthquake indicates epidemiological changes attributable to the event itself.
Earlier reports on extreme violence (armed groups, sexual and domestic abuse) and frequent traffic accidents among the Port-au-Prince population help to explain why injury seems even more present in years without a disaster 1,16. The unexpected decrease in violent and accidental trauma in the aftermath of the earthquake is probably, respectively, because of a low profile of prostrated armed groups and the fact that roads were impassable.
Possible explanations for the marked occurrence of respiratory, ophthalmological, digestive and skin disorders following the earthquake compared with other years may be atomization of rubble dust, poor hygiene circumstances in huddled IDP camps and increased awareness of respiratory symptoms following the 2009 Haitian Ministry of Health H1/N1 flu campaign 15,17.
The increased proportions of postevent neurological and psychological diagnoses are attributable to headaches because of stress and disrupted sleep from nightmares and anxiety, which corresponds with earlier documentation on psychological trauma following earthquakes 20.
No significant changes in the occurrence of genitourinary and circulatory diseases were observed in the postearthquake period.
Almost 38% of all patients examined by B-FAST presented with infectious diseases in the 4-week period after the earthquake. This is almost twice the baseline observed in the same period of earlier and later years. A peak incidence occurred as of the third week after the earthquake: 88% of the children under 5 and almost 45% of older patients suffered from infections.
The B-FAST findings on the marked proportion of respiratory, digestive and ophthalmological diseases are consistent with earlier reports on diagnoses after earthquakes and other disasters 13,21–23. This study emphasizes that an important share of these are indeed earthquake related, and not only governed by seasonal variation or IDP huddle and poor sanitary circumstances.
Limitations and strengths
This study has several limitations. B-FAST and MSF worked under different circumstances and settings. Although both teams used forms based on WHO health cards and defined diagnoses from the same manual, the lack of uniform standards to register complaints, clinical features and diagnoses made comparison of datasets challenging 6,7,17. Final diagnoses remain tentative because of a lack of lab and imaging equipment. Exclusion of patients because of missing data (in carrying out research while working against time under complex circumstances, obviously the patient has priority) has been reported by other FMT 5,7,8,14. This study covered a limited number of Port-au-Prince districts, preventing extrapolation of results to other areas. Estimating the total population of Delmas and Martissant regions at 658,513, the B-FAST and MSF cohort samples each represent less than 1% of this population 24.
The strengths of this study are the considerable number of patients and the comparison of diagnoses, made by physicians, of before and after an earthquake, never documented in this way before 6.
Interpretation and conclusion
For FMT to meet the changing needs of victims in subsequent phases of the disaster response, the first to be sent to the affected area are surgeons, anaesthesiologists, intensivists and ancillary nurses, together with the necessary equipment and supplies 25. There is an early role for emergency physicians, paediatricians, gynaecologists, midwives and pharmacists 23,26. FMTs should scale up within 2 weeks with internists, rehabilitation teams, psychologists and public healthcare personnel 2,8,26.
The existing IEHKs should be adapted to the specific medical needs of disaster victims in this phase by providing respiratory, digestive and ophthalmological drugs and supplies, rehabilitation materials, as well as paediatric formulas of essential medicines 3,15,26.
More research is needed to confirm the important share of medical problems in victims and IDP camps during the aftermath of different types of disasters.
More epidemiological data are needed to adapt and improve the composition of FMTs and the content of IEHKs to be deployed for all types of disasters 3,26.
The development of a standardized template, to prospectively register and subsequently report health data as well before as after disasters, would allow relevant research to improve disaster preparedness, management and mitigation on the basis of evidence in the future 4,7–9,12,18,27. This would imply that healthcare workers locally active in disaster risk areas use the same standards to register baseline data in daily medical practice as FMT during emergency relief 14,18,28,29.
As we cannot prevent these disasters from occurring, we can at least attempt to be maximally informed and ideally prepared to deal with (most of) them.
Conflicts of interest
There are no conflicts of interest.
1. Kolbe AR, Hutson RA, Shannon H, Trzcinski E, Miles B, Levitz N, et al. Mortality, crime and access to basic needs before and after the Haiti earthquake: a random survey of Port-au-Prince households. Med Confl Surviv 2010; 26:281–297.
2. Gerdin M, Wladis A, von Schreeb J. Foreign field hospitals after the 2010 Haiti earthquake: how good were we? Emerg Med J 2013; 30:e8.
3. World Health Organisation. The Interagency Emergency Health Kit 2011. World Health Organ Tech Rep Ser 2011. [Epub ahead of print].
4. Debacker M, Hubloue I, Dhondt E, Rockenschaub G, Rüter A, Codreanu T, et al. Utstein-style template for uniform data reporting of acute medical response in disasters. PLoS Curr 2012; 4.
5. World Medical Association. WMA declaration of Helsinki – ethical principles for medical research involving human subjects. Bull World Health Organ 2001; 79:373–374.
6. Auf der Heide E. The importance of evidence-based disaster planning. Ann Emerg Med 2006; 47:34–49.
7. Redmond AD, Mardel S, Taithe B, Calvot T, Gosney J, Duttine A, Girois S. A qualitative and quantitative study of the surgical and rehabilitation response to the earthquake in Haiti, January 2010. Prehosp Disaster Med 2011; 26:449–456.
8. Gerdin M, Clarke M, Allen C, Kayabu B, Summerskill W, Devane D, et al. Optimal evidence in difficult settings: improving health interventions and decision making in disasters. PLoS Med 2014; 11:e1001632.
9. Kar-Purkayastha I, Clarke M, Murray V. Dealing with disaster databases – what can we learn from health and systematic reviews? Application in practice. PLoS Curr 2011; 3:RRN1272.
10. Doocy S, Daniels A, Packer C, Dick A, Kirsch T. The human impact of earthquakes: a historical review of events 1980-2009 and systematic literature review. PLoS Curr 2013; 5.
11. Kreiss Y, Merin O, Peleg K, Levy G, Vinker S, Sagi R, et al. Early disaster response in Haiti: the Israeli field hospital experience. Ann Intern Med 2010; 153:45–48.
12. Burnweit C, Stylianos S. Disaster response in a pediatric field hospital: lessons learned in Haiti. J Pediatr Surg 2011; 46:1131–1139.
13. Bartels SA, VanRooyen MJ. Medical complications associated with earthquakes. Lancet 2012; 379:748–757.
14. Johnson J, Galea S. Cherry KE. Disasters and population health. Lifespan perspectives on natural disasters. Dordrecht, Heidelberg, London, New York: Springer; 2009. 281–326.
15. Noji EK. The public health consequences of disasters. Prehosp Disaster Med 2000; 15:147–157.
16. Polonsky J, Luquero F, Francois G, Rousseau C, Caleo G, Ciglenecki I, et al. Public health surveillance after the 2010 haiti earthquake: the experience of médecins sans frontières. PLoS Curr 2013; 5.
17. Connolly MA. Communicable disease control in emergencies: a field manual. Geneva: World Health Organization; 2005. 204–231.
18. Bambrick AT, Passman DB, Torman RM, Livinski AA, Olsen JM. Optimizing the use of chief complaint & diagnosis for operational decision making: an EMR case study of the 2010 Haiti earthquake. PLoS Curr 2014; 6.
19. Newcombe R. Two-sided confidence intervals for the single proportion: comparison of seven methods. Stat Med 1998; 17:857–872.
20. North CS, Pfefferbaum B. Mental health response to community disasters: a systematic review. JAMA 2013; 310:507–518.
21. Roggen I, van Berlaer G, Gijs G, Hubloue I. Clinical characteristics of the inhabitants of an internally displaced persons camp in Brazzaville, Republic of Congo after the arms dump blast on March 4, 2012. Prehosp Disaster Med 2014; 29:516–520.
22. Chen KT, Chen WJ, Malilay J, Twu SJ. The public health response to the Chi-Chi earthquake in Taiwan, 1999. Public Health Rep 2003; 118:493–499.
23. Broach JP, McNamara M, Harrison K. Ambulatory care by disaster responders in the tent camps of Port-au-Prince, Haiti, January 2010. Disaster Med Public Health Prep 2010; 4:116–121.
24. Haitian Institute for Statistics and Information IHSI. Haiti in numbers. Available at: http://www.ihsi.ht/haiti_en_chiffre.htm
[Accessed 25 January 2016].
25. Norton I, von Schreeb J, Aitken P, Herard P, Lajolo C. Classification and minimum standards for foreign medical teams in sudden onset disasters. Geneva: World Health Organisation; 2013.
26. World Health Organization. Manual for the health care of children in humanitarian emergencies. Geneva: World Health Organization; 2008.
27. Jafar AJ, Norton I, Lecky F, Redmond AD. A literature review of medical record keeping by foreign medical teams in sudden onset disasters. Prehosp Disaster Med 2015; 30:216–222.
28. Evidence Aid Priority Setting Group EA. Prioritization of themes and research questions for health outcomes in natural disasters, humanitarian crises or other major healthcare emergencies. PLoS Curr 2013; 5.
29. Burkle FM Jr, Nickerson JW, von Schreeb J, Redmond AD, McQueen KA, Norton I, Roy N. Emergency surgery data and documentation reporting forms for sudden-onset humanitarian crises, natural disasters and the existing burden of surgical disease. Prehosp Disaster Med 2012; 27:577–582.