Impact of local circumstances on outcome of renal casualties in major disasters

Transcription

1 Nephrol Dial Transplant (2009) 24: doi: /ndt/gfn557 Advance Access publication 8 October 2008 Original Article Impact of local circumstances on outcome of renal casualties in major disasters Arjan van der Tol 1, Asrar Hussain 2, Memhet Sukru Sever 3, Stefaan Claus 1, Wim Van Biesen 1, Eric Hoste 4, Samuel Khan 2 and Raymond Vanholder 1,5 1 Renal Division, Department of Internal Medicine, University Hospital, Ghent, Belgium, 2 Department of Nephrology, Pakistan Institute of Medical Sciences, Islamabad, Pakistan, 3 Department of Internal Medicine, Istanbul School of Medicine, Istanbul, Turkey, 4 Intensive Care Unit 1K12-C, University Hospital, Ghent and 5 Chairman Renal Disaster Relief Task Force, International Society of Nephrology, Brussels, Belgium Abstract Background. In the aftermath of earthquakes, the cumulative incidence of crush-induced acute kidney injury (AKI) is difficult to predict. Insight into factors determining this risk is indispensable to allow adequate logistical planning, which is a prerogative for success in disaster management. Methods. Data of 88 crush-related AKI patients in the aftermath of the Kashmir earthquake were collected and outcome measures were analysed. Then the findings were compared with the data of 596 crush-related AKI patients of the Marmara earthquake. Results. The earthquake in Kashmir occurred in a rural area with lack of medical facilities and difficult transportation conditions while the earthquake in Marmara occurred in an urban area with more efficient transport possibilities. In Kashmir we reported fewer patients with treated AKI (1.2 AKI per 1000 deaths, 1.3 AKI per 1000 victims) than in Marmara (34.1 AKI per 1000 deaths; P < 0.001, 13.6 AKI per 1000 victims; P < 0.001). Time lag between earthquake and admission to hospitals was longer in Kashmir (5.8 ± 5.8 days) than in Marmara (3.5 ± 3.7 days; P < 0.001). The frequencies of fasciotomies (P < 0.001), amputations (P < 0.001) and dialysis (P = 0.005) were lower in Kashmir, than in Marmara AKI patients. Conclusions. The cumulative incidence of treated AKI related to number of deaths or victims might differ substantially among earthquakes. Many factors may affect the frequency of AKI: hampered rescue and transport possibilities; destroyed medical facilities on the spot; availability or not of sophisticated therapeutic possibilities and structure of the buildings might all have impacted on different cumulative incidence between Kashmir and Marmara. Keywords: acute kidney injury; crush syndrome; earthquakes; Kashmir; Marmara Correspondence and offprint requests to: Arjan van der Tol, Renal Division, Department of internal Medicine, University Hospital, De Pintelaan 185, 9000 Ghent, Belgium. Tel: ; Fax: ; arjan.vandertol@ugent.be Introduction Highly vulnerably built environments that cannot withstand the energy input from earthquakes cause high mortality and morbidity [1]. The higher the intensity on the Richter scale of an earthquake, the higher the death to injury ratio [2]. Most earthquake fatalities die as a result of asphyxia/compression or head/neck trauma within the first hours after the quake [3,4]. Surviving victims who are entrapped in collapsed buildings are at risk for crush syndrome [5]. The crush syndrome is caused by rhabdomyolysis due to compression and ischaemic injury to the skeletal muscles [6]. Fluid therapy prevents renal failure in crush victims that is induced by the presence of dehydration and acidosis together with release of myoglobulin into the circulation [7]. The cumulative incidence of actual AKI in surviving victims in the aftermath of disasters is difficult to detect for practical reasons. The treated cumulative incidence of AKI shows remarkable variability among various disasters and is affected by rescue efforts, entrapment time, timing of the disaster and prehospital patient dehydration [8]. For example, the Indian Gujarat earthquake in 2001 caused deaths, injured and only 35 treated AKI [9] while the Japanese Hanshin earthquake in 1995 caused fewer deaths (5500) and injured (41 000) but 202 treated AKI [10]. In the present study, the cumulative incidence of treated AKI and its relation to the local circumstances are compared between the Kashmir earthquake on 8 October 2005 and the Marmara earthquake in Northwestern Turkey on 17 August Methods The Renal Disaster Relief Task Force (RDRTF) of the International Society of Nephrology (ISN) offered support for the renal problems in the aftermath of the Kashmir and Marmara earthquakes, next to other events [11]. In Kashmir C The Author [2008]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 908 A. van der Tol et al. the data were collected from the files of 93 crush victims with renal problems who were admitted to renal wards or intensive care units from five different Pakistani hospitals: Pakistan Institute of Medical Sciences (n = 43); SHIFA international (n = 11); CGH Rawalpindi (n = 14); Military Hospital-Rawalpindi (n = 11) and Fanji Foundation Hospital Rawalpindi (n = 9). A local Pakistani doctor (A Hussain) filled out the questionnaires in the aftermath of the disaster. In Marmara the data were also collected by questionnaires that were sent to 35 different nephrology units that treated victims with crush-induced renal problems. Most patients were first admitted to the emergency unit and were then referred to the renal unit after clinical and biochemistry examination. The local physicians decided upon the indication for dialysis. Because all the patients presented with acute renal dysfunction without any baseline measurement of renal function, we calculated a theoretical baseline serum creatinine value in each patient of the study, assuming that a simplified modification of diet in renal disease MDRD formula provides a robust estimate of a normal GFR ( ml/min per 1.73 m 2 ) relative to serum creatinine based on age, race and sex. The change in serum creatinine was estimated in each patient. AKI was defined according to the RIFLE criteria as an abrupt reduction in kidney function corresponding to an absolute at least twofold increase in serum creatinine or/and urine output <60 ml/kg/12 u [12,13]. In Kashmir five patients were not included in the database because they did not fulfil the RIFLE criteria or because of missing important data. According to the RIFLE criteria, 596 patients had AKI of the 639 patients with renal problems due to crush injury in the Marmara cohort. We combined the data of the patients with RIFLE- AKI from both earthquakes in one SPSS data file. Statistical analysis We recorded the data in SPSS 15.0 Descriptive statistics for all numeric variables, including means, medians, standard deviations and minimum and maximum values. We performed the t-test to compare two independent group means. When the continuous variables were not normally distributed, the Mann Whitney U-test was used. The proportions of all categorical variables were calculated in 2 2 contingency tables; differences between group proportions were examined by the χ 2 test. The calculated P-values were provided. A binominal test was used to compare the gender distribution of the patients with AKI to the gender ratio in the local population (test proportion). Results In this study we compared various features, which could be associated with the risk of AKI in the aftermath of the Kashmir and Marmara earthquake that both occurred in densely populated and inhabited regions. The intensity of the earthquake was slightly higher in Kashmir than in Marmara (Table 1). The Kashmir disaster caused more total deaths, more injured survivors and a higher death to injury ratio than the Marmara earthquake [1] (Table 1), but Table 1. Local circumstances and the outcomes in the Kashmir and in the Marmara disasters Kashmir 2005 Marmara 1999 Richter scale Number of AKI Immediate seismic deaths a Injured survivors a Ratio deaths/injured 1:1 2:5 P < AKI/death P < AKI/injured victims P < Dialyzed AKI/death P < AKI cases/million P < inhabitants Total population a b Urban inhabitants a c P < Rural inhabitants a c P < Inhabitants per km a 250 c Cumulative incidence of AKI P < a &id=2257&itemid=144. b Sever MS[18]; figures related to Istanbul are not included because only a small part of the city was affected by the earthquake. c = 235. in Kashmir fewer patients with treated crush-induced AKI (n = 88) were reported than in Marmara (n = 596). The earthquake in Kashmir hit in the morning while the earthquake in Marmara took place at night. In the Kashmir earthquake more than children died and many were injured in collapsed poorly built school buildings [14,15]. Both AKI groups had a low number of children probably because of lower surface area and/or muscle mass than adults. The proportions of the age categories (Figure 1A) were not significantly different between both AKI populations although the global population in Kashmir contained more children ( ) than in Marmara ( ; Figure 1B). The mean age of AKI victims in Kashmir was 35.7 ± 14.8 years while in Marmara they were slightly younger (32.2 ± 14.4; P = 0.03 [16]). More males than females suffered from AKI in both populations. In Kashmir the male to female ratio in the victims with AKI was 1.32 while the gender ratio in the global population was This difference was not significant, probably due to the small numbers in the AKI group (n = 88). In Marmara the male to female ratio in the renal victims was 1.23 while the gender ratio in the global population was [17,18]. This difference was significant (P = 0.03). The victims with AKI in Kashmir were admitted markedly later than the AKI victims in the urban areas of Marmara (Table 2; Figure 2; P < 0.001). The mean serum creatinine, phosphorus and calcium in AKI victims were higher in Kashmir than in Marmara (Table 3). The peak CK level in AKI victims in Marmara was higher than in Kashmir (Table 3). The frequencies of fasciotomies, amputations and the need for renal replacement therapy were markedly lower in the Kashmir patients than in Marmara (Table 2). We reported more fractures of the extremities in the Kashmir AKI patients than in Marmara (P < 0.001). The Marmara patients with AKI had more thorax traumas than the Kashmir victims with AKI (P < 0.001). The

3 Impact of local circumstances on outcome of renal casualties 909 Fig. 1. (A) The proportion (%) of the local population by age categories in Kashmir (dark bars) and Marmara [1] (grey bars), P < (B) The proportion (%) of the patients with AKI by age categories in Kashmir (dark bars) and Marmara (grey bars), P = NS. mortality in both AKI groups was not significantly different (Table 2). The majority of the population in Azad Jammu, Kashmir and the North West Frontier Province [19] lived in rural areas with many adobe or brick masonry buildings [20] while the majority of the people in the east of the Marmara Sea lived in urban areas with many multi-storey concrete buildings [21]. The victims with AKI in Kashmir were entrapped for a shorter time than in Marmara (Table 2). In Kashmir a higher proportion of patients with AKI were entrapped less than 4 h than in Marmara (Table 2). In Marmara a higher proportion of patients with AKI were entrapped >24 h than in Kashmir (Table 2). Discussion To our knowledge, until now there have been no direct studies comparing different earthquake conditions regarding their relation to renal outcome. It is of utmost importance to compare outcomes of disasters allowing us to make more accurate predictions in future events. In the present study we evaluated the influence of demographical and geological features in two major earthquakes of the last decade on the cumulative incidence of treated crush-induced AKI. Admittedly, such an analysis can never be the topic of a controlled study, due to disaster circumstances, and can only be pursued by comparison of events with different characteristics. The high altitude and unstable mountainous geological situation make Northern Pakistan very vulnerable to extensive landslides, especially when an earthquake strikes the region [22]. This phenomenon made the Kashmir region less accessible and delayed the rescue efforts, resulting in a high risk to die on the spot, hence indirectly reducing the cumulative incidence of treated AKI. The importance of fast rescue was well illustrated in the Kobe disaster that revealed a high survival rate of the extricated victims on

4 910 A. van der Tol et al. Table 2. Time period under the rubble and before admission, major complications, interventions and outcome in AKI patients in Kashmir and in Marmara Kashmir Marmara P Table 3. Metabolic profile on admission in patients with crush-induced AKI in Kashmir and Marmara Laboratory Kashmir Marmara P results (n = 88) (n = 596) Time period under the 7.2 ± 6.1 n = ± 13.3 n = 500 <0.001 rubble (hours) Time stratum 0 3 a 19 (22%) 26 (5%) <0.001 Time stratum 4 24 a 66 (77%) 437 (88%) Time stratum >25 a 1(1%) 34(7%) 0.04 Time period before 5.8 ± ± 3.7 <0.001 admission (d b ) Fasciotomy 22/88 (26%) 312/596 (52%) <0.001 Amputation 3/88 (3%) 93/596 (16%) <0.001 Dialysis 55/88 (63%) 475/596 (80%) <0.001 Fractures of the 52/88 (59%) 122/596 (20%) <0.001 extremities Abdominal trauma 6/88 (7%) 39/596 (11%) 0.91 Thorax trauma 3/88 (3%) 65/596 (11%) <0.001 Outcome (mortality) 16/85 (19%) 98/497 (20%) 0.94 a Total number of victims with AKI (%) according to time strata of entrapment under the rubble in hours. b Time delay in days between the occurrence of the earthquake and admission to hospitals. the first day and a low probability of being alive on extrication after the first day [23]. Further possible explanation for the lower number of treated AKI could be the selection bias of patients during the triage process, whereby those needing dialysis or intensive care may have been excluded from transport, because of the awareness of low availability of both amenities. Patients with less severe injuries might have been given higher priority. Also some patients or their relatives refused to seek medical help because of economic or religious reasons. The crush victims with AKI in Marmara had more crushinduced complications and necessitated more interventions (dialysis, fasciotomies and amputations) while the AKI patients in Kashmir had more fractures of the extremities; this may be related to the daytime occurrence of the earthquake in Kashmir when most individuals are upright and about, whereas the Marmara earthquake occurred overnight and Serum creatinine 6.4 ± 4.8 n = ± 2.9 n = Serum 5.4 ± 1.1 n = ± 1.3 n = potassium (meq/l) Serum CK (U/l) 9,400 ± n = 81 15,300 ± n = ,000 27,600 Serum 6.9 ± 2.8 n = ± 2.1 n = 323 <0.001 phosphorus Serum calcium 8.3 ± 1.3 n = ± 1.8 n = 433 <0.001 Serum 2.9 ± 0.6 n = ± 1.8 n = albumin hit people in supine position that increased the risk for complex muscle compression [24,25]. Another reason for fewer crush-induced complications in the Kashmir group could be related to the shorter time period under the rubble than the Marmara group of AKI who were entrapped in more complex buildings. The time period under the rubble is pointed out as an important indicator for crush-induced AKI [26 30]. The Bam earthquake (2003, 5.30 a.m.) showed that crush-related AKI (n = 170) developed in victims who were entrapped longer whereas victims developing no AKI were extricated earlier (6.2 ± 4.1 versus 2.1 ± 3.9 h; P < 0.001) [31]. Further important factors besides speed of rescue efforts, entrapment period and time of occurrence that influence the frequency of crush-induced complications during disasters are the degree of availability of basic life supplies and the meteorological circumstances. For example, the victims in Marmara were more prone to dehydration because the quake occurred in a hot summer Fig. 2. Time lag between the occurrence of the earthquake and admission to hospitals in days (d): Percentage (%) of victims with AKI in Kashmir (dark bars) and Marmara (grey bars), P <

5 Impact of local circumstances on outcome of renal casualties 911 Table 4. Renal Disaster Relief Task Force crush syndrome patients questionnaire 1. Demography Name Age Gender Chronic illness City (of origin) City (where Date of Time under before disaster treated) admission rubble 2. Trauma Extremity trauma Abdominal trauma Thoracic trauma Other 3. Admission findings Blood pressure First 24 h urine volume First urine colour CK, K, Crea, BUN Hb, WBC, Plt Ca, P Uric acid 4. RIFLE criteria R = Risk of renal dysfunction I = injury to the kidney F = failure of kidney function L = loss of kidney function E = end-stage kidney disease 5. Interventions At admission/during clinical course Surgical interventions Fasciotomy/amputation Medical interventions Antibiotics period, which therefore very likely caused more crushinduced AKI. Our findings were later on confirmed by our observations in the aftermath of the Yogyakarta earthquake in Indonesia, 27 May 2006, which occurred early in the morning and in an essentially rural area, with light building structures, and easy rescue [personal observation at the occasion of an intervention from the Renal Disaster Relief Task Force (RDRTF) by S. Claus and A. van der Tol]. The earthquake had an intensity of 6.3 on Richter scale and caused 5782 deaths and injured [32], but only four cases of registered AKI needing dialysis. Most surviving earthquake victims were extricated quickly and could be offered intravenous fluid, both preventing crush-induced AKI, because the Yogyakarta area was already prepared for a disaster due to the activity of the Merapi volcano. Also, the experience gained in the aftermath of the latest Peruvian earthquake, which with an intensity of 7.9 took place in the evening of 15 August 2007 indicates that local conditions have a major impact on the frequency of AKI. Even for the maximum reported mortality rate of 519 and 1366 injured [33], which remained unmodified from Day 3 on, the occurrence of one single AKI victim needing dialysis is low. Here the late start of rescue activities (reportedly at the earliest 14 h after the disaster), partly due not only to the vast destruction of roads but also to a lack of advance rescue plans, together with the light building structures and the daytime occurrence, led to an extremely low number of polytrauma victims in the hospitals and to a low incidence of crush-induced AKI [personal observation at the occasion of an intervention from the Renal Disaster Relief Task Force (RDRTF) by S. Claus]. Our study suffers from a number of unavoidable shortcomings since collecting data during chaotic disaster conditions will inevitably lead to a database with missing values. Our data are not adequate enough to offer an exact quantification of the actual cumulative incidence of AKI because a serum creatinine check could not be performed in the large majority of surviving victims. In addition the patients who had their AKI diagnosed may have been a selected group and in this way not typical of the majority of AKI patients in Kashmir, as a large part may have died before reaching facilities where diagnosis was possible. Destroyed medical facilities on the spot and delayed transfers of the severely injured victims affected the high mortality and deteriorated the outcome of the surviving injured resulting in less treated AKI. In future disaster planning, it will be a challenge to screen the patients with AKI and to transfer the surviving injured casualties to hospitals with modern facilities, to close the gap between actual AKI and treated AKI, especially in rural areas with fewer medical supplies. The Renal Disaster Relief Task Force should adapt their support to these local circumstances. To identify patients with AKI in chaotic conditions, a crush syndrome patients questionnaire could be very helpful. In Table 4 we illustrate the simple questionnaire that is currently used by RDRTF. Also, a protocol for fluid administration to prevent AKI, distribution of intravenous fluids and a dry-biochemistry kit in (rural) affected areas are needed in future earthquakes. Conflict of interest statement. None declared.

6 912 A. van der Tol et al. References 1. Sever MS, Vanholder R, Lameire N. Management of crush-related injuries after disasters. NEJM 2006; 354: Gutiérrez E, Taucer F, De Groeve T et al. Analysis of worldwide earthquake mortality using multivariate demographic and seismic data. Am J Epidemiol 2005; 161: Chan CC, Lin YP, Chen HH et al. A population-based study on the immediate and prolonged effects of the 1999 Taiwan earthquake on mortality. Ann Epidemiol 2003; 13: Aoki N, Nishimura A, Pretto EA et al. Survival and cost analysis of fatalities of the Kobe earthquake in Japan. Prehosp Emerg Care 2004; 8: Vanholder R, Van Der Tol A, Sever MS et al. Earthquake and crush syndrome casualties: lessons learned from the Kashmir disaster. Kidney Int 2007; 71: Yokota J. Crush syndrome in disaster. JMAJ 2005; 48: Gunal AI, Celiker H, Dogukan A et al. Early and vigorous fluid resuscitation prevents acute renal failure in the crush victims of catastrophic earthquakes. J Am Soc Nephrol 2004; 15: Sheng ZY. Medical support in the Tangshan earthquake: a review of the management of mass casualties and certain major injuries. J Trauma 1987; 27: Viroja D et al. Management of crush syndrome following Gujarat earthquake. Nephrol Dial Transplant 2001; W357(suppl 4): Oda J, Tanaka H, Yoshioka T et al. Analysis of 372 patients with crush syndrome caused by the Hanshin-Awaji earthquake. J Trauma 1997; 42: Vanholder R, Van Biesen W, Lameire N et al. International Society of Nephrology/Renal Disaster Relief Task Force: the role of the International Society of Nephrology/Renal Disaster Relief Task Force in the rescue of renal disaster victims. Contrib Nephrol 2007;156: Kellum JA, Bellomo R, Ronco C. The concept of acute kidney injury and the Rifle criteria. Contrib Nephrol 2007; 156: Bellomo R, RoncoC, Kellum J et al. and the ADQI workgroup. Acute renal failure definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004; 8: R204 R Helminen M, Saarela E, Salmela J. Characterization of patients treated at the Red Cross field hospital in Kashmir during the first three weeks of operation. Emerg Med J 2006; 23: Laverick S et al. Asian earthquake: report from the first volunteer British hospital team in Pakistan. Emerg Med J 2007; 24: Sever MS. The Crush Syndrome. First Ed. Basel: S Karger, 2005, Sever MS. The Crush Syndrome. First Ed. Basel: S Karger, 2005, Peiris N, Rossetto T, Burton P et al. Earthquake Engineering Field Investigation Team: Mission: October 8, 2005 Kashmir Earthquake Tanaka K. The Kobe earthquake: the system response. A disaster report from Japan. Eur J Emerg Med 1996; 3: Roy N, Shan H, Patel V et al. The Gujarat earthquake (2001) experience in a seismically unprepared area: community hospital medical response. Prehospital Disaster Med 2002; 17: Mahue-Giangreco M, Mack W, Seliason H et al. Risk factors associated with moderate and serious injuries attributable to the 1994 Northridge earthquake, LA, California. Ann Epidemiol 2001; 11: Collins AJ, Burzstein S. Renal failure in disasters. Crit Care Clin 1991; 7: Better OS. Acute renal failure in casualties of mass disasters. Kidney Int 1993; 43(suppl 41): Atef MR, Nadjaffi I, BoroumandBet al. Acute renal failure in earthquake victims in Iran: epidemiology and management. QJMed1994; 87: Sever MS, Erek E, Vanholder R et al. Lessons learned from the Marmara disaster: time period under the rubble. Crit Care Med 2002; 30: Sever MS, Erek E, Vanholder R et al. The Marmara earthquake: epidemiological analysis of the victims with nephrological problems. Kidney Int 2001; 60: Hatamizadeh P, Najafi I, Vanholder R et al. Epidemiologic aspects of the Bam earthquake in Iran: the nephrologic perspective. Am J Kidney Dis 2006; 47: Received for publication: Accepted in revised form: