Advertisement

Risk factors for severe malaria among hospitalized patients in the United States, 2000–2014

Published:February 14, 2018DOI:https://doi.org/10.1016/j.idh.2018.01.002

      Highlights

      • There were an estimated 4823 severe malaria-related hospitalizations from 2000 to 2014 in the US.
      • Pf accounted for most of the severe malaria-related hospitalizations, followed by Pv, Po, and Pm.
      • Severe malaria was most common among inpatients who were male, Black, and aged 45–64 years.
      • We found associations with severe malaria or specific complications for age, sex, and HIV.

      Abstract

      Background

      Factors associated with the development of severe malaria have not been well described for cases occurring in the United States (US).

      Methods

      Severe malaria hospitalizations data from the 2000–2014 Nationwide Inpatient Sample were analyzed. Frequencies were reported by demographic, clinical, species, financial, geographic, and institutional characteristics, and trends and disparities were identified. Logistic regression models were used to identify potential predictors for severe disease among those with malaria.

      Results

      From 2000 to 2014, there were an estimated 4823 severe malaria cases, representing 21.9% of all malaria-related hospitalizations, including 182 severe malaria deaths. Severe malaria was most common among inpatients who were male, Black, aged 45–64 years, and hospitalized in the South Atlantic division of the US. Older age was associated with higher odds of severe malaria, cerebral malaria, ARDS, severe anemia, and renal failure. Males had higher odds of developing renal failure and jaundice, while females had higher odds of developing severe anemia. HIV infection was associated with increased odds of severe malaria, severe anemia, and renal failure.

      Conclusion

      Primary and secondary prevention measures, such as pre-travel consultations, chemoprophylaxis, and early diagnosis and treatment, should be emphasized and improved among high-risk prospective travelers to malaria endemic countries.

      Keywords

      Introduction

      Malaria is the leading cause of death by parasitic disease in the world and remains one of the most important and intractable global public health problems. The WHO estimated that 212 million cases of malaria, and 429,000 deaths due to malaria occurred in 2015 [
      • World Malaria Report 2016
      World health organization.
      ], though others have estimated that the number of deaths may be much higher [
      • Murray C.J.L.
      • Ortblad K.F.
      • Guinovart C.
      • Lim S.S.
      • Wolock T.M.
      • Roberts D.A.
      • et al.
      Global, regional, and national incidence and mortality for HIV, tuberculosis, and malaria during 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013.
      ]. Malaria is caused by infection with the protozoan agents of the genus Plasmodium. Several species of Plasmodium (P. falciparum, P. vivax, P. ovale, P. malariae, and P. knowlesi) are known to affect humans, with P. falciparum causing the most morbidity and the vast majority of the mortality. Transmission typically occurs through the bite from an infective female Anopheles mosquito. These competent vectors have widespread distribution throughout the world, including the United States [

      Centers for Disease Control. Malaria. Available at: http://www.cdc.gov/malaria [Accessed 15 April 2017].

      ].
      Clinical manifestations can range from fever, shaking chills, muscle pains, and other non-specific symptoms in uncomplicated malaria, to jaundice, acute renal failure, severe anemia, cerebral malaria, acute respiratory distress syndrome, and other serious complications in severe malaria, which can be rapidly fatal. Malaria during pregnancy is associated with many adverse outcomes, including maternal mortality, maternal anemia, low birth weight, intrauterine growth retardation, and fetal loss. Malaria can progress to severe and fatal disease, even with prompt treatment [

      Centers for Disease Control. Malaria. Available at: http://www.cdc.gov/malaria [Accessed 15 April 2017].

      ].
      The CDC publishes an annual malaria report that discusses reported cases of malaria in the US. This surveillance data indicates that the number of imported uncomplicated malaria and severe malaria cases has steadily increased over the last few decades in the US [
      • Cullen K.A.
      • Mace K.E.
      • Arguin P.M.
      Malaria surveillance–United States, 2013.
      ] and in other countries [
      • Checkley A.M.
      • Smith A.
      • Smith V.
      • Blaze M.
      • Bradley D.
      • Chiodini P.L.
      • et al.
      Risk factors for mortality from imported falciparum malaria in the United Kingdom over 20 years: an observational study.
      ] where previously endemic malaria has been eliminated. These cases mostly occur among returned travelers, and to a lesser extent, among foreign visitors or immigrants, from malaria-endemic countries [
      • Hagmann S.H.F.
      • Van P.V.
      • Stauffer W.M.
      • Miller A.O.
      • Connor B.A.
      • Hale D.C.
      • et al.
      Travel-associated disease among US residents visiting US GeoSentinel clinics after return from international travel.
      ]. Some studies have suggested the risk of uncomplicated and severe malaria is unevenly distributed across different subpopulations of travelers [
      • Checkley A.M.
      • Smith A.
      • Smith V.
      • Blaze M.
      • Bradley D.
      • Chiodini P.L.
      • et al.
      Risk factors for mortality from imported falciparum malaria in the United Kingdom over 20 years: an observational study.
      ,
      • Odolini S.
      • Gautret P.
      • Parola P.
      Epidemiology of imported malaria in the Mediterranean region.
      ,
      • Seringe E.
      • Thellier M.
      • Fontanet A.
      • Legros F.
      • Bouchaud O.
      • Ancelle T.
      • et al.
      Severe imported Plasmodium falciparum malaria, France, 1996–2003.
      ]. However, despite the large disease and economic burden [
      • Khuu D.
      • Eberhard M.L.
      • Bristow B.N.
      • Javanbakht M.
      • Ash L.R.
      • Shafir S.C.
      • et al.
      Malaria-related hospitalizations in the United States, 2000-2014.
      ], the factors associated with the development of severe disease due to malaria infection have not been well described for cases occurring in the US. A stratified analysis of the specific malaria complications has rarely been conducted, and is of interest since each complication may have different risk factors, require different types and levels of resources and expertise, and result in different outcomes. In the current study, we analyzed US hospital data to examine the patient characteristics and to identify potential risk factors for severe malaria, which could be used to target interventions for high-risk travelers to improve health care or the clinical outcomes of malaria in the US. This paper focused on severe malaria follows our previous paper on the overall malaria hospitalization burden [
      • Khuu D.
      • Eberhard M.L.
      • Bristow B.N.
      • Javanbakht M.
      • Ash L.R.
      • Shafir S.C.
      • et al.
      Malaria-related hospitalizations in the United States, 2000-2014.
      ].

      Methods

      Hospital discharge records from the Nationwide Inpatient Sample (NIS) were used for analysis of severe malaria-related hospitalizations in the US during 2000–2014. The NIS is sampled from the State Inpatient Databases (SID), and is part of the Healthcare Cost and Utilization Project (HCUP) sponsored by the Agency for Healthcare Research and Quality (AHRQ). The NIS includes data on sociodemographics, diagnosis type, disease severity, length of stay, co-diagnoses, institutional characteristics, and total charges. Details on the sampling scheme have been described elsewhere [
      • Healthcare Cost and Utilization Project (HCUP)
      Introduction to the HCUP Nationwide Inpatient Sample (NIS) 2011.
      ,
      • Healthcare Cost and Utilization Project (HCUP)
      Introduction to the HCUP Nationwide inpatient sample (NIS) 2012.
      ].
      Cases of malaria in the NIS from 2000 to 2014 were identified from discharge records in the NIS by the primary and secondary diagnoses, which used the International Classification of Diseases, 9th revision (ICD-9) codes of 084.0–084.9 (084.0: falciparum malaria, malignant tertian; 084.1: vivax malaria, benign tertian; 084.2: malariae, quartan; 084.3: ovale malaria; 084.4: other malaria; 084.5: mixed malaria; 084.6: malaria, unspecified; 084.7: induced malaria; 084.8: Blackwater fever; 084.9: other pernicious complications of malaria) and of 647.4 (malaria complicating pregnancy, childbirth, or puerperium) [

      Centers for Disease Control and Prevention. Classification of diseases, functioning, and disability. Conversion table. Available at: http://www.cdc.gov/nchs/icd/icd9cm.htm. [Accessed 31 March 2017].

      ].
      In this study, the definition of severe malaria was modified from that used by the CDC [
      • Cullen K.A.
      • Mace K.E.
      • Arguin P.M.
      Malaria surveillance–United States, 2013.
      ]. Since no specific drugs used or laboratory results are available in the NIS, we were unable to use parasitemia ≥5% or treatment for severe malaria (i.e. artesunate or quinidine) as specified in the CDC definition [
      • Checkley A.M.
      • Smith A.
      • Smith V.
      • Blaze M.
      • Bradley D.
      • Chiodini P.L.
      • et al.
      Risk factors for mortality from imported falciparum malaria in the United Kingdom over 20 years: an observational study.
      ]. Malaria complications were identified using ICD-9 codes and HCUP Clinical Classification Software (CCS) categories, which are clinically meaningful categories of ICD-9 codes [
      • HCUP Clinical Classifications Software
      Healthcare cost and utilization Project (HCUP).
      ]. Discharge records listing a malaria diagnosis along with one or more of the following criteria were considered as severe malaria cases: 1. Neurologic symptoms (cerebral malaria)– CCS codes 82, 83, 85, or 95 (paralysis, epilepsy, convulsions, alteration of consciousness, coma, stupor, brain damage, other nervous system disorders); 2. Renal failure– CCS code 157; 3. Severe anemia– CCS codes 59 (deficiency anemia), with procedural CCS code 222 (blood transfusion); 4. Acute respiratory distress syndrome (ARDS)– CCS code 131; 5. Jaundice– ICD-9 code 782.4; or 6. Exchange transfusion– ICD-9 code 99.01. Malaria-related hospitalizations with an in-hospital death were also considered as severe malaria cases.
      All-Patient Refined Diagnosis Related Groups (APR-DRG) severity and risk of mortality variables, which measure the extent of physiologic decompensation or organ system loss of function, and the likelihood of death (based on the combinations of principal diagnoses, secondary diagnoses, non-operating room procedures, operating room procedures, and age) [
      • 3M APR DRG Classification System
      Methodology overview.
      ], were additionally used to describe the uncomplicated malaria and severe malaria cases.
      Data analyses were performed using SAS 9.4 (SAS Institute, Cary, North Carolina, USA) with survey procedures that accounted for the complex sampling design. National estimates of the frequency by different characteristics were produced using NIS discharge-level sample weights, which are provided by HCUP. Rates per 1 million US population were calculated using bridged-race US census population estimates [

      United States Department of Health and Human Services (US DHHS), Centers for disease control and prevention (CDC), national center for health statistics (NCHS), bridged-race population estimates, United States July 1st resident population by state, county, age, sex, bridged-race, and hispanic origin. Compiled from 1990-1999 bridged-race intercensal population estimates (released by NCHS on 7/26/2004); revised bridged-race 2000-2009 intercensal population estimates (released by NCHS on 10/26/2012); and bridged-race Vintage 2014 (2010-2014) postcensal population estimates (released by NCHS on 6/30/2015). Available on CDC WONDER Online Database. Accessed at http://wonder.cdc.gov/bridged-race-v2014.html on Apr 15, 2016 1:00:23 PM

      ]. Trends in the number of severe malaria hospitalizations over the study period were assessed by negative binomial regression offset by the population.
      Separate multiple logistic regression models were used to identify potential predictors for severe malaria, and for each severe malaria defining complication among malaria-related hospitalizations in the US from 2000 to 2014. Variables and covariates were selected based on associations documented in the literature, which included sociodemographic, financial, and institutional characteristics, and some pre-existing conditions, such as HIV [
      • Chalwe V.
      • Van geertruyden Jean-Pierre
      • Mukwamataba D.
      • Menten J.
      • Kamalamba J.
      • Mulenga M.
      • et al.
      Increased risk for severe malaria in HIV-1–infected adults, Zambia.
      ], type II diabetes mellitus [
      • Danquah I.
      • Bedu-Addo G.
      • Mockenhaupt F.P.
      Type 2 diabetes mellitus and increased risk for malaria infection.
      ], and essential hypertension [
      • Gallego-Delgado J.
      • Rodriguez A.
      Malaria and hypertension. Another co-evolutionary adaptation?.
      ].

      Results

      From 2000 to 2014, there were 22,029 (95% confidence interval [CI]: 21,001–23,057) malaria-related hospitalizations, of which an estimated 4823 (95% CI: 4485–5161) were severe malaria hospitalizations (Table 1, Fig. 1), and 182 (95% CI: 123–241) were severe malaria deaths. Severe malaria hospitalizations occurred at a rate of 1.07 (95% CI: 0.99–1.14) per 1 million population (Table 2, Fig. 1). Severe malaria accounted for 21.9% of all malaria-related hospitalizations in the US over the study period, with the percent of all malaria-related hospitalizations that were severe cases increasing from 11.2% in 2000 to 34.4% in 2014. The rate of severe malaria overall (ptrend ≤ 0.05, Fig. 1) and of each specific malaria complication (all ptrend ≤ 0.05), except death, increased over the study period.
      Table 1Demographics of severe malaria hospitalizations in the US, by complication, 2000–2014.
      Uncomplicated malariaSevere malariaMalaria with ARDSCerebral malariaMalaria with severe anemiaMalaria with renal failureMalaria with jaundiceFatal malaria
      N(%)N(%)N(%)N(%)N(%)N(%)N(%)N(%)
      Sex
       Male10,189(59.2)3055(63.3)505(56.1)600(62.8)736(46.4)1587(75.1)633(78.4)111(61.0)
       Female6930(40.3)1754(36.4)395(43.9)351(36.7)842(53.1)527(24.9)174(21.6)71(39.1)
      Pregnant1113(6.5)107(2.2)10(1.1)
      Small cells.
      88(5.2)
      Small cells.
      Small cells.
      Race
       White3300(19.2)1153(23.9)315(35)326(34.1)311(19.6)460(21.8)154(19.0)54(29.6)
       Black7501(43.6)2234(46.3)289(32.1)416(43.5)706(44.5)1112(52.6)386(47.8)63(34.8)
       Hispanic978(5.7)191(4.0)28(3.1)36(3.8)93(5.9)64(3.0)33(4.1)
      Small cells.
       Asian/Pacific  Islander888(5.2)215(4.5)50(5.6)44(4.6)104(6.6)89(4.2)12(1.5)15(8.2)
       Other1691(9.8)396(8.2)87(9.7)53(5.6)181(11.4)137(6.5)89(11.0)20(10.7)
      Age (years)
       Under 152189(12.7)399(8.3)46(5.1)69(7.2)216(13.6)64(3.0)69(8.5)
      Small cells.
       15–243115(18.1)445(9.2)79(8.7)61(6.4)165(10.4)140(6.6)113(14.0)
      Small cells.
       25–446663(38.7)1485(30.8)232(25.8)203(21.2)477(30.1)603(28.5)333(41.2)30(16.2)
       45–644104(23.9)1715(35.6)329(36.6)401(41.9)515(32.4)921(43.6)263(32.6)73(40.1)
       65 or older1105(6.4)770(16.0)214(23.8)218(22.8)209(13.2)386(18.2)30(3.7)64(35.1)
       Mean35.143.948.348.740.548.738.655.6
      Total (% of severe malaria)17,206(78.1)
      Percent of all malaria hospitalizations.
      4823(21.9)
      Percent of all malaria hospitalizations.
      900(18.7)956(19.8)1587(32.9)2113(43.8)808(16.8)182(3.8)
      Note: All numbers are national estimates based on weighted frequencies. Numbers and proportions may not sum to total or 100% due to rounding and missing values.
      a Small cells.
      b Percent of all malaria hospitalizations.
      Figure thumbnail gr1
      Figure 1Number of severe malaria hospitalizations in the US, by year, 2000–2014. Note: All numbers are national estimates based on weighted frequencies.
      Table 2Rates
      Rates per 1 million US population.
      of severe malaria in the US, by demographic, NIS, 2000–2014.
      CharacteristicUncomplicated malariaSevere malariaMalaria with ARDSCerebral malariaMalaria with severe anemiaMalaria with renal failureMalaria with jaundiceFatal malaria
      Rate(95% CI)Rate(95% CI)Rate(95% CI)Rate(95% CI)Rate(95% CI)Rate(95% CI)Rate(95% CI)Rate(95% CI)
      Sex
       Male4.59(4.33, 4.86)1.38(1.26, 1.50)0.23(0.18, 0.27)0.27(0.18, 0.27)0.33(0.28, 0.39)0.72(0.63, 0.80)0.290.290.05(0.03, 0.07)
       Female3.02(2.81, 3.22)0.76(0.68, 0.85)0.17(0.13, 0.21)0.17(0.13, 0.21)0.37(0.31, 0.42)0.23(0.19, 0.27)0.080.080.03(0.01, 0.05)
      Race
       White1.10(1.01, 1.20)0.39(0.34, 0.44)0.11(0.08, 0.13)0.11(0.08, 0.14)0.10(0.08, 0.13)0.15(0.12, 0.19)0.050.050.02(0.01, 0.03)
       Black13.08(12.00, 14.17)3.90(3.48, 4.32)0.50(0.37, 0.63)0.73(0.56, 0.89)1.23(1.02, 1.44)1.94(1.67, 2.21)0.670.670.11(0.05, 0.17)
       Hispanic1.42(1.18, 1.66)0.28(0.19, 0.37)0.04(0.01, 0.07)0.05(0.02, 0.09)0.14(0.08, 0.20)0.09(0.04, 0.14)0.050.05
      Small cells.
       API11.55(10.41, 12.69)2.74(2.22, 3.25)0.61(0.38, 0.85)0.43(0.24, 0.62)1.28(0.94, 1.62)1.01(0.71, 1.31)0.450.450.15(0.04, 0.27)
      Age (years)
       Under 142.40(2.07, 2.73)0.44(0.34, 0.54)0.05(0.02, 0.08)0.08(0.02, 0.08)0.24(0.16, 0.31)0.07(0.03, 0.11)0.080.080.01(0.00, 0.02)
       15–244.88(4.44, 5.31)0.70(0.55, 0.84)0.12(0.06, 0.18)0.10(0.06, 0.18)0.26(0.17, 0.35)0.22(0.14, 0.30)0.180.180.01(0.00, 0.04)
       25–445.34(4.99, 5.69)1.19(1.05, 1.33)0.19(0.13, 0.24)0.16(0.13, 0.24)0.38(0.31, 0.46)0.48(0.39, 0.57)0.270.270.02(0.00, 0.04)
       45–643.63(3.33, 3.92)1.52(1.34, 1.69)0.29(0.22, 0.36)0.35(0.22, 0.36)0.45(0.37, 0.54)0.81(0.69, 0.93)0.230.230.06(0.03, 0.10)
       65 or  older1.89(1.64, 2.14)1.32(1.11, 1.53)0.37(0.26, 0.48)0.37(0.26, 0.48)0.36(0.25, 0.46)0.66(0.51, 0.81)0.050.050.11(0.05, 0.17)
      Total3.81(3.62, 4.00)1.07(0.99, 1.14)0.20(0.17, 0.23)0.21(0.18, 0.24)0.35(0.31, 0.39)0.47(0.42, 0.51)0.180.180.04(0.03, 0.05)
      Note: All numbers are national estimates based on weighted frequencies. Numbers and proportions may not sum to total or 100% due to missing values.
      API: Asian and Pacific Islander.
      a Small cells.
      b Rates per 1 million US population.
      Some (25.6%) patients developed multiple complications. The most frequent malaria complication was renal failure, followed by severe anemia, cerebral malaria, ARDS, jaundice, and death (Table 1, Table 2). Those with ARDS had the highest risk of death (13.1%), accounting for 64.8% of all severe malaria-related hospitalizations ending in death (Table 2). Controlling for demographic covariates (age, sex, and race), having ARDS (OR = 32.25, 95% CI: 14.62–71.16), renal failure (OR = 9.57, 95% CI: 4.24–21.62), cerebral malaria (OR = 7.00, 95% CI: 2.90–16.85), or severe anemia (OR = 2.97, 95% CI: 1.22–7.21) was associated with increased odds of death.

      Species

      Infecting Plasmodium species information was known for 52.9% of severe malaria-related hospitalizations. While P. falciparum accounted for 69.7% of uncomplicated malaria-related hospitalizations, it accounted for 83.3% of severe malaria-related hospitalizations with known species information, followed by P. vivax (13.8%), P. ovale (2.4%), and P. malariae (1.5%). A similar pattern held for all of the severe malaria complications (Fig. 2). Temporal increases in the rates of severe malaria for infections with P. falciparum (ptrend ≤ 0.01) were observed.
      Figure thumbnail gr2
      Figure 2Uncomplicated and severe malaria hospitalizations in the US, by infecting species, 2000–2014. Note: Based on malaria-related hospitalizations with known infecting species. Totals may exceed 100% due to mixed infections.

      Demographics

      The number of and rate severe malaria-related hospitalizations, and each complication except jaundice, increased with age, peaking at 45–64 years (Table 1, Table 2). Older age was associated with severe malaria, malaria with renal failure, ARDS, cerebral malaria, severe anemia, and fatal malaria (Table 3). Temporal increases in the rates of severe malaria were observed for all age groups (ptrend ≤ 0.01).
      Table 3Multiple logistic regression analysis of severe malaria in the US, by complication, 2000–2014.
      CovariateOutcome
      Severe malariaMalaria with ARDSCerebral malariaMalaria with severe anemiaMalaria with renal failureMalaria with jaundice
      OR(95% CI)OR(95% CI)OR(95% CI)OR(95% CI)OR(95% CI)OR(95% CI)
      Age (ref: age 0–14)
       Age 15–240.74(0.51, 1.08)1.08(0.43, 2.67)0.52(0.22, 1.21)0.57(0.34, 0.97)1.83(0.85, 3.96)0.97(0.44, 2.16)
       Age 25–441.15(0.84, 1.57)1.50(0.67, 3.36)0.82(0.42, 1.60)0.79(0.51, 1.21)2.95(1.51, 5.79)1.39(0.72, 2.67)
       Age 45–642.15(1.55, 2.98)2.97(1.33, 6.63)2.20(1.14, 4.26)1.44(0.91, 2.27)6.64(3.34, 13.22)1.84(0.94, 3.59)
       Age 65+4.61(2.87, 7.39)5.76(2.13, 15.55)2.89(1.05, 7.98)2.89(1.14, 5.94)16.26(7.17, 36.88)0.65(0.26, 1.59)
      Sex (ref: male)
       Female0.87(0.73, 1.04)1.41(1.00, 1.98)0.96(0.70, 1.33)1.97(1.52, 2.55)0.46(0.36, 0.60)0.32(0.19, 0.52)
      Race (ref: white)
       Black0.92(0.74, 1.14)0.56(0.33, 0.83)0.70(0.46, 1.07)1.02(0.72, 1.45)1.19(0.87, 1.62)1.17(0.70, 1.94)
       Hispanic0.51(0.34, 0.76)0.33(0.13, 0.86)0.49(0.22, 1.09)1.14(0.64, 2.03)0.45(0.23, 0.89)0.33(0.09, 1.16)
       Other race0.72(0.55, 0.96)0.76(0.45, 1.29)0.49(0.28, 0.84)1.26(0.83, 1.93)0.74(0.49, 1.11)0.91(0.48, 1.70)
      Comorbidity (ref: no)
       HIV2.35(1.32, 4.18)1.56(0.47, 5.18)2.16(0.81, 5.73)2.62(1.27, 5.43)2.80(1.46, 5.38)0.27(0.04, 1.94)
       Diabetes0.95(0.71, 1.28)1.06(0.60, 1.88)0.93(0.55, 1.57)1.16(0.75, 1.80)1.20(0.82, 1.74)0.83(0.36, 1.92)
       Hypertension1.07(0.84, 1.35)1.15(0.74, 1.77)1.14(0.75, 1.72)0.66(0.44, 1.01)1.24(0.91, 1.68)0.64(0.34, 1.20)
      Income (ref: high)
       Low1.17(0.98, 1.38)0.95(0.67, 1.33)1.08(0.77, 1.53)1.18(0.91, 1.54)1.24(0.97, 1.59)1.01(0.68, 1.50)
      Payer (ref: Medicare)
       Medicaid1.33(0.83, 2.15)1.04(0.43, 2.49)0.94(0.36, 2.43)2.31(1.04, 5.13)1.88(1.03, 3.43)1.33(0.55, 3.18)
       Private insurance1.19(0.76, 1.85)0.91(0.42, 1.96)0.65(0.27, 1.56)2.04(0.94, 4.42)1.71(0.99, 2.96)1.11(0.52, 2.36)
       Self-pay1.39(0.88, 2.21)1.00(0.44, 2.25)0.64(0.25, 1.64)2.09(0.94, 4.65)2.30(1.30, 4.08)1.32(0.59, 2.98)
       No charge1.13(0.57, 2.24)0.30(0.04, 2.68)0.63(0.16, 2.57)1.68(0.53, 5.31)1.66(0.64, 4.31)0.84(0.17, 4.01)
       Other payer2.17(1.23, 3.83)2.31(0.91, 5.89)1.10(0.38, 3.24)2.94(1.17, 7.38)2.41(1.13, 5.14)1.93(0.74, 5.04)
      Region (ref: west)
       Northeast0.61(0.47, 0.80)0.54(0.32, 0.91)0.49(0.30, 0.81)0.74(0.48, 1.13)0.71(0.48, 1.05)0.65(0.36, 1.19)
       Midwest0.66(0.46, 0.96)0.55(0.26, 1.16)0.51(0.25, 1.06)0.65(0.34, 1.22)1.07(0.63, 1.80)0.84(0.37, 1.93)
       South1.07(0.83, 1.38)0.89(0.56, 1.41)0.71(0.46, 1.11)1.14(0.77, 1.70)1.10(0.76, 1.58)1.05(0.61, 1.81)
      Location (ref: urban)
       Rural0.55(0.34, 0.90)0.78(0.33, 1.83)0.64(0.27, 1.52)0.60(0.24, 1.48)0.19(0.06, 0.60)0.43(0.10, 1.89)
      Severe malaria-related hospitalization generally occurred more often among males, and the rate disparity was greatest for those with jaundice and renal failure. Malaria with severe anemia was an exception, and was more common among females (Table 1, Table 2). Of the 107 (6.1% of female cases) severe malaria hospitalizations among pregnant women, 88 (82.2%) had severe anemia (Table 1). Males had higher odds of developing renal failure and jaundice, while females had higher odds of developing severe anemia (Table 3). Temporal increases in the rates of severe malaria for both males and females were observed (ptrend ≤ 0.01).
      Blacks were the most common race group, accounting for over half (53.3%) of severe malaria-related hospitalizations with known race information, followed by Whites, Asian/Pacific Islanders, and Hispanics (Table 1). Blacks also had the highest rates of severe malaria overall compared to the other race groups (Table 2). Temporal increases in the rates of severe malaria for Whites, Blacks, and Asian/Pacific Islanders (ptrend ≤ 0.01) were observed. Whites had nearly twice (OR = 1.96, 95% CI: 1.32–2.94) the odds of developing severe malaria compared to Hispanics after controlling for all other covariates (Table 3).

      Clinical

      Malaria was the primary diagnosis (73.7%) for the majority of severe malaria-related hospitalizations, though the proportion varied widely by complication. Those with jaundice (91.4%) were most likely, while those with cerebral malaria (54.6%) were least likely to have a primary malaria diagnosis (Table 4). The most common primary diagnoses among severe malaria cases with a secondary malaria diagnosis were unspecified septicemia (21.0%) and other specified septicemia (6.5%).
      Table 4Clinical characteristics of severe malaria hospitalizations in the US, by complication, 2000–2014.
      Uncomplicated malariaSevere malariaMalaria with ARDSCerebral malariaMalaria with severe anemiaMalaria with renal failureMalaria with jaundiceFatal malaria
      N(%)N(%)N(%)N(%)N(%)N(%)N(%)N(%)
      Malaria diagnosis
       Primary14,740(85.7)3557(73.7)570(63.3)522(54.6)1221(76.9)1591(75.3)739(91.4)80(44.2)
       Secondary2466(14.3)1266(26.3)330(36.7)434(45.4)367(23.1)522(24.7)69(8.5)102(55.9)
      Pre-existing conditions
       HIV205(1.2)133(2.8)23(2.5)26(2.7)47(3.0)84(4.0)
      Small cells (<10) were suppressed to preserve confidentiality.
      14(7.7)
       Diabetes  (type II)1163(6.8)501(10.4)110(12.2)107(11.2)161(10.2)301(14.3)48(5.9)15(8.2)
       Essential  hypertension2134(12.4)985(20.4)210(23.3)226(23.7)230(14.5)562(26.6)93(11.6)43(23.5)
      Procedures performed
       013,042(75.8)1704(35.3)125(13.9)320(33.4)
      Small cells (<10) were suppressed to preserve confidentiality.
      929(43.9)487(60.3)29(16.2)
       12538(14.7)1297(26.9)75(8.3)175(18.3)717(45.2)415(19.6)158(19.6)4(2.3)
       2 or more1626(9.5)1822(37.8)700(77.8)461(48.2)871(54.9)770(36.4)162(20.0)148(81.6)
      Severity of illness
      All patient refined diagnosis related group (APRDRG) severity of illness and risk of mortality based on 2002–2014 data only.
       No/minor  loss of function3944(28.1)94(2.1)
      Small cells (<10) were suppressed to preserve confidentiality.
      29(3.3)61(4.2)
      Small cells (<10) were suppressed to preserve confidentiality.
      Small cells (<10) were suppressed to preserve confidentiality.
      Small cells (<10) were suppressed to preserve confidentiality.
       Moderate  loss of function7247(51.6)891(20.2)
      Small cells (<10) were suppressed to preserve confidentiality.
      199(22.6)353(24.2)100(5.0)255(34.8)
      Small cells (<10) were suppressed to preserve confidentiality.
       Major loss  of function2644(18.8)2160(48.9)155(19.3)281(31.9)645(44.3)1085(54.0)275(53.1)39(24.7)
       Extreme  loss of function130(0.9)1269(28.7)638(79.4)368(41.8)398(27.3)820(40.8)64(12.2)113(72.5)
      Risk of mortality
      All patient refined diagnosis related group (APRDRG) severity of illness and risk of mortality based on 2002–2014 data only.
       No/minor  risk of dying10,806(77.0)997(22.6)
      Small cells (<10) were suppressed to preserve confidentiality.
      200(22.8)387(26.6)30(1.5)370(50.4)
      Small cells (<10) were suppressed to preserve confidentiality.
       Moderate  risk of dying2470(17.6)1255(28.4)29(3.6)185(21.0)421(28.9)619(30.8)212(28.8)
      Small cells (<10) were suppressed to preserve confidentiality.
       Major risk  of dying614(4.4)1221(27.6)215(26.8)196(22.3)349(23.9)736(36.7)74(10.1)45(28.8)
       Extreme  risk of dying77(0.5)941(21.3)549(68.3)294(33.4)299(20.5)619(30.8)79(10.8)107(68.4)
       Mean length  of stay (days)3.47.816.010.28.58.44.611.8
       In-hospital death0(0.0)182(3.8)118(13.1)55(5.8)41(2.6)106(5.0)
      Small cells (<10) were suppressed to preserve confidentiality.
      182(100.0)
      Total (% of severe malaria)17,206(78.1)
      Percent of all malaria hospitalizations.
      4823(21.9)
      Percent of all malaria hospitalizations.
      900(18.7)956(19.8)1587(32.9)2113(43.8)808(16.8)182(3.8)
      Note: All numbers are national estimates based on weighted frequencies. Numbers and proportions may not sum to total or 100% due to rounding and missing values.
      a Small cells (<10) were suppressed to preserve confidentiality.
      b All patient refined diagnosis related group (APRDRG) severity of illness and risk of mortality based on 2002–2014 data only.
      c Percent of all malaria hospitalizations.
      HIV infection was reported in 2.8%, type II diabetes in 10.4%, and essential hypertension in 20.4% of severe malaria hospitalizations (Table 4). Controlling for all other covariates, HIV infection increased the odds of severe malaria, malaria with severe anemia, and malaria with renal failure. Type II diabetes and essential hypertension were not associated with severe malaria or any severe malaria complication (Table 3).
      Most (77.6%) patients with severe malaria had major or extreme physiologic decompensation or organ system loss of function, and about half (48.9%) had a major or extreme risk of dying. Nearly all patients with ARDS had major or extreme loss of organ system function (98.7%) and major or extreme risk of dying (95.1%) (Table 4).
      The mean length of stay for severe malaria-related hospitalizations was 7.8 days. Those with ARDS had the longest hospital stays (16.0 days), while those with jaundice had the shortest hospital stays (4.6 days) on average (Table 4).

      Financial

      Severe malaria-related hospitalizations accounted for $276,657,680 in hospital charges from 2000 to 2014, with a mean hospital charge of $58,500. The mean charge varied greatly among complications, ranging from a mean of $32,087 for malaria with jaundice, to $155,873 for malaria with ARDS (Table 5). Malaria patients who were self-payers or had Medicaid were at increased odds of having renal failure, and those who had Medicaid were at increased odds of having severe anemia, compared to those with Medicare (Table 3) (Note: Medicaid is a publicly funded health insurance program for low-income and needy people. Medicare is a publicly funded health insurance program for people who are 65 or older, certain younger people with disabilities, and people with end-stage renal disease).
      Table 5Financial characteristics of severe malaria hospitalizations in the US, by complication, 2000–2014.
      Uncomplicated malariaSevere malariaMalaria with ARDSCerebral malariaMalaria with severe anemiaMalaria with renal failureMalaria with jaundiceFatal malaria
      N(%)N(%)N(%)N(%)N(%)N(%)N(%)N(%)
      Mean charge ($)
      Adjusted for inflation to 2015 US dollars.
      16,58658,500155,87379,13062,07466,86932,087108,825
      Income
      Income category is based on median household income by national quartiles for patient zip code.
       High9398(54.6)2557(53.0)536(59.5)525(54.9)831(52.4)1092(51.7)414(51.3)122(67.3)
       Low6792(39.5)1950(40.4)332(36.8)391(40.9)652(41.1)886(41.9)313(38.7)55(30.2)
      Primary payer
       Medicare948(5.5)568(11.8)169(18.8)191(20.0)124(7.8)257(12.1)25(3.1)30(16.5)
       Medicaid3419(19.9)841(17.4)124(13.7)158(16.5)360(22.7)313(14.8)141(17.5)35(19.5)
       Private insurance7977(46.4)1960(40.6)371(41.2)363(37.9)660(41.6)887(42)345(42.7)68(37.5)
       Self-pay3745(21.8)1017(21.1)145(16.1)159(16.7)318(20.1)489(23.1)213(26.4)44(23.9)
       No charge348(2.0)103(2.1)
      Small cells (<10) were suppressed to preserve confidentiality.
      21(2.2)29(1.8)48(2.3)14(1.8)
      Small cells (<10) were suppressed to preserve confidentiality.
       Other728(4.2)316(6.5)82(9.1)59(6.2)91(5.7)115(5.4)69(8.5)
      Small cells (<10) were suppressed to preserve confidentiality.
      Total (% of severe malaria)17,206(78.1)
      Percent of all malaria hospitalizations.
      4823(21.9)
      Percent of all malaria hospitalizations.
      900(18.7)956(19.8)1587(32.9)2113(43.8)808(16.8)182(3.8)
      Note: All numbers are national estimates based on weighted frequencies. Numbers and proportions may not sum to total or 100% due to rounding and missing values.
      a Small cells (<10) were suppressed to preserve confidentiality.
      b Adjusted for inflation to 2015 US dollars.
      c Income category is based on median household income by national quartiles for patient zip code.
      d Percent of all malaria hospitalizations.

      Institutional

      The South Atlantic division of the US had the highest numbers of severe malaria-related hospitalizations, followed by the Mid-Atlantic, Pacific, West South Central, East North Central, New England, Mountain, West North Central, and East South Central divisions (Table 6). Controlling for all other covariates, those in the Northeastern and Midwestern region were associated with lower odds of severe malaria, compared to the patients treated in the Western region of the US. Those in the Northeastern region had lower odds of cerebral malaria and ARDS (Table 3).
      Table 6Institutional characteristics of severe malaria hospitalizations in the US, by complication, 2000–2014.
      CharacteristicUncomplicated malariaSevere malariaMalaria with ARDSCerebral malariaMalaria with severe anemiaMalaria with renal failureMalaria with jaundiceFatal malaria
      N(%)N(%)N(%)N(%)N(%)N(%)N(%)N(%)
      Region/Division
      New England: ME, NH, VT, MA, RI, CT; Mid-Atlantic: NY, PA, NJ; East North Central: WI, MI, IL, IN, OH; West North Central: MO, ND, SD, NE, KS, MN, IA; South Atlantic: DE, MD, DC, VA, WV, NC, SC, GA, FL; East South Central: KY, TN, MS, AL; West South Central: OK, TX, AR, LA; Mountain: ID, MT, WY, NV, UT, CO, AZ, NM; Pacific: AK, WA, OR, CA, HI.
       Northeast6137(35.7)1386(28.7)210(23.3)256(26.8)481(30.3)640(30.3)223(27.6)59(32.4)
      New England1239(7.2)312(6.5)62(6.9)72(7.5)130(8.2)121(5.7)40(5.1)11(6.2)
      Mid-Atlantic4898(28.6)1074(22.4)148(16.5)185(19.3)351(22.2)519(24.6)183(22.9)48(26.1)
       Midwest2536(14.7)596(12.3)124(13.8)91(9.5)161(10.1)273(12.9)119(14.7)35(19.3)
      East North Central1739(10.1)433(9.0)78(8.7)76(7.9)107(6.8)194(9.2)94(11.7)30(16.6)
      West North Central797(4.6)162(3.4)46(5.2)15(1.6)54(3.4)79(3.7)25(3.1)
      Small cells (<10) were suppressed to preserve confidentiality.
       South6074(35.3)2089(43.3)381(42.4)390(40.8)690(43.5)923(43.7)340(42.1)58(31.6)
      South Atlantic4069(23.7)1392(29.0)265(29.6)301(31.5)433(27.4)618(29.2)224(28.1)39(21.3)
      East South Central309(1.8)150(3.1)33(3.6)15(1.6)33(2.1)53(2.5)44(5.6)
      Small cells (<10) were suppressed to preserve confidentiality.
      West South Central1627(9.5)539(11.2)80(8.9)74(7.7)224(14.2)252(11.9)68(8.5)14(7.8)
       West2459(14.3)752(15.6)185(20.6)219(22.9)256(16.1)278(13.1)126(15.6)31(16.8)
      Mountain660(3.9)197(3.9)53(6.0)60(6.3)66(4.2)83(3.9)23(2.9)10(5.6)
      Pacific1798(10.5)543(10.5)132(14.7)158(16.6)184(11.6)195(9.2)97(12.1)20(11.2)
      Hospital Location
       Rural677(3.9)134(2.8)34(3.7)27(2.8)45(2.9)20(0.9)18(2.3)
      Small cells (<10) were suppressed to preserve confidentiality.
       Urban16,506(95.9)4674(96.9)861(95.7)923(96.6)1542(97.2)2083(98.6)785(97.1)182(100.0)
      Total (% of severe malaria)17,206(78.1)
      Percent of all malaria hospitalizations.
      4823(21.9)
      Percent of all malaria hospitalizations.
      900(18.7)956(19.8)1587(32.9)2113(43.8)808(16.8)182(3.8)
      Note: All numbers are national estimates based on weighted frequencies. Numbers and proportions may not sum to total or 100% due to rounding and missing values.
      a Small cells (<10) were suppressed to preserve confidentiality.
      b New England: ME, NH, VT, MA, RI, CT; Mid-Atlantic: NY, PA, NJ; East North Central: WI, MI, IL, IN, OH; West North Central: MO, ND, SD, NE, KS, MN, IA; South Atlantic: DE, MD, DC, VA, WV, NC, SC, GA, FL; East South Central: KY, TN, MS, AL; West South Central: OK, TX, AR, LA; Mountain: ID, MT, WY, NV, UT, CO, AZ, NM; Pacific: AK, WA, OR, CA, HI.
      c Percent of all malaria hospitalizations.
      Patients with a severe malaria diagnosis were more often hospitalized in urban areas (95.9%) (Table 6). Rural hospital location was associated with lower odds of severe malaria, particularly for renal failure (Table 3).

      Discussion

      Severe and fatal malaria exacts a considerable, and increasing, burden in the US. This burden may be greater than previously recognized, as the number and proportion of severe malaria cases (and deaths) of all malaria cases identified from the NIS was higher than those reported in the malaria surveillance system. For example, in 2013, there were an estimated 400 (30.8%) severe malaria-related hospitalizations, compared to 270 (15.6%) reported severe cases [
      • Cullen K.A.
      • Mace K.E.
      • Arguin P.M.
      Malaria surveillance–United States, 2013.
      ]. This increased number of observed cases is expected, given underreporting in passive surveillance systems [
      • Abanyie F.A.
      • Arguin P.M.
      • Gutman J.
      State of malaria diagnostic testing at clinical laboratories in the United States, 2010: a nationwide survey.
      ,
      • Doyle T.J.
      • Glynn M.K.
      • Groseclose S.L.
      Completeness of notifiable infectious disease reporting in the United States: an analytical literature review.
      ]. In addition, as previously noted elsewhere [
      • Das B.S.
      Renal failure in malaria.
      ], the number of complications and co-diagnoses of severe malaria cases has gradually increased in recent years, perhaps in part due to the general increase in more complete reporting of co-diagnoses over the years.
      Based on cases with known species information, the predominant infecting species for severe malaria-related hospitalizations was P. falciparum, which is consistent with its well-documented virulence [
      • Walker D.M.
      • Oghumu S.
      • Gupta G.
      • McGwire B.S.
      • Drew M.E.
      • Satoskar A.R.
      Mechanisms of cellular invasion by intracellular parasites.
      ]. However, the other species also produced serious disease classifiable as severe malaria, and should not be discounted [
      • Hwang J.
      • Cullen K.A.
      • Kachur S.P.
      • Arguin P.M.
      • Baird J.K.
      Severe morbidity and mortality risk from malaria in the United States, 1985–2011.
      ,
      • Mendis K.
      • Sina B.J.
      • Marchesini P.
      • Carter R.
      The neglected burden of Plasmodium vivax malaria.
      ]. In particular, although cerebral malaria is a severe clinical manifestation of malaria traditionally associated almost exclusively with P. falciparum [
      • Trampuz A.
      • Jereb M.
      • Muzlovic I.
      • Prabhu R.M.
      Clinical review: severe malaria.
      ], 17.7% of cerebral malaria hospitalizations were attributed to P. vivax mono-infection in this study. This adds to the growing body of evidence that P. vivax may cause serious disease including cerebral malaria [
      • Kochar D.K.
      • Saxena V.
      • Singh N.
      • Kochar S.K.
      • Kumar S.V.
      • Das A.
      Plasmodium vivax malaria.
      ], which may be especially relevant for diagnosis of returning travelers from P. vivax-endemic Latin America or Asia presenting with fever and neurological symptoms. Additional laboratory training on species identification and quantification through malaria microscopy [
      • Ash L.R.
      • Orihel T.C.
      Atlas of human parasitology.
      ] may be needed, as only about half of severe malaria hospitalizations had species information. The CDC currently recommends species confirmation by PCR after microscopic diagnosis for all malaria cases in the US [

      Centers for Disease Control. Malaria. Available at: http://www.cdc.gov/malaria [Accessed 15 April 2017].

      ].
      Older age was found to increase the risk of ARDS, cerebral malaria, severe anemia, and renal failure; male sex was found to increase the risk of renal failure and jaundice; and female sex was found to increase the risk of severe anemia among patients with malaria in this study, which is consistent with previous reports [
      • Mühlberger N.
      • Jelinek T.
      • Behrens R.H.
      • Gjørup I.
      • Coulaud P.
      • Clerinx J.
      • et al.
      Age as a risk factor for severe manifestations and fatal outcome of falciparum malaria in European patients: observations from TropNetEurop and SIMPID Surveillance Data.
      ,
      • Leder K.
      • Black J.
      • O’Brien D.
      • Greenwood Z.
      • Kain K.C.
      • Schwartz E.
      • et al.
      Malaria in travelers: a review of the GeoSentinel surveillance network.
      ,
      • Coggins C.H.
      • Breyer Lewis J.
      • Caggiula A.W.
      • Castaldo L.S.
      • Klahr S.
      • Wang S.R.
      Differences between women and men with chronic renal disease.
      ,
      • Yu Z.
      • Zhan J.
      • Li C.
      • Zhou H.
      Age and gender analysis of jaundice patients.
      ,
      • Käser A.K.
      • Arguin P.M.
      • Chiodini P.L.
      • Smith V.
      • Delmont J.
      • Jimenez B.C.
      • et al.
      Imported malaria in pregnant women: a retrospective pooled analysis.
      ]. Although we observed an association between race and severe malaria, race may serve as a proxy for other risk factors that may better explain the variation in Plasmodium infection and the subsequent development of severe malaria. Such factors may include infecting species and level of malaria transmission at the travel destinations, patient travel history, chemoprophylaxis use, purpose of travel, genetic profiles, underlying medical conditions, and perceived immune status.
      A large proportion of severe malaria cases did not list malaria as the primary diagnosis, particularly among fatal and cerebral malaria cases, and a large proportion listed unspecified or other septicemia as the primary diagnosis. This demonstrates the non-specific nature of malarial illness and a potentially substantial lack of clinical suspicion, which may be addressed with increased awareness and training on malaria among physicians.
      The proportion of severe malaria patients with HIV infection (2.8%) was much higher than that of the hospitalized (0.6%) and general (0.3%) [
      • Hall H.I.
      • An Q.
      • Tang T.
      • Song R.
      • Chen M.
      • Green T.
      • et al.
      Prevalence of diagnosed and undiagnosed HIV infection – United States, 2008–2012.
      ] populations. Having HIV infection as a diagnosis was associated with increased odds of severe malaria in this study, which is consistent with the findings of other studies [
      • Chalwe V.
      • Van geertruyden Jean-Pierre
      • Mukwamataba D.
      • Menten J.
      • Kamalamba J.
      • Mulenga M.
      • et al.
      Increased risk for severe malaria in HIV-1–infected adults, Zambia.
      ]. HIV infection may compromise the immune response to Plasmodium, resulting in higher parasitemia and more severe disease.
      The proportion of severe malaria patients with type II diabetes (10.4%) was lower than that of the hospitalized population (15.8%) and similar to that of the general population (9.3%) [
      • Centers for Disease Control and Prevention
      National diabetes statistics report: estimates of diabetes and its burden in the United States, 2014.
      ], while the proportion of severe malaria patients with essential hypertension (20.4%) was lower than that of the hospitalized (31.9%) and general (29.1%) [
      • Nwankwo T.
      • Yoon S.S.
      • Burt V.
      • Gu Q.
      Hypertension among adults in the United States: national health and nutrition examination survey, 2011–2012. NCHS data brief, no 133.
      ] populations. Some studies have suggested that type II diabetes may be a risk factor [
      • Danquah I.
      • Bedu-Addo G.
      • Mockenhaupt F.P.
      Type 2 diabetes mellitus and increased risk for malaria infection.
      ], and that hypertension may be a protective factor for severe malaria [
      • Gallego-Delgado J.
      • Rodriguez A.
      Malaria and hypertension. Another co-evolutionary adaptation?.
      ]. However, we did not find type II diabetes or essential hypertension to be associated with severe malaria or any specific complication of malaria.
      Severe malaria hospitalizations accounted for a large and unnecessary economic burden, which could largely be prevented through use of effective personal protective measures against mosquitoes and adherence to proper chemoprophylaxis. Although chemoprophylaxis is often not covered by insurance, which could deter to its use, the cost disparity warrants consideration among prospective travelers.
      Severe malaria cases may be referred from smaller, rural hospitals to larger urban hospitals, which tend to offer a higher level of care. This may explain the observed lower odds of renal failure at rural hospitals. Differences in medical practice patterns or unmeasured patient characteristics may have led to differences in recognition and management of malaria by region.

      Limitations

      The use of NIS in this study has many strengths, including its large and representative sample size to study rare diseases. However, some limitations of this study should be considered. The national estimates from the NIS using ICD codes are subject to coding and sampling errors, and the actual number of severe malaria-related hospitalizations may be different from that reported here. Underestimates may result from misdiagnoses, in which malaria may not have been considered as a diagnosis and a travel history was not elicited due to disease rarity, and low patient, clinician, and laboratory personnel experience with malaria in the US. Without information on specific medications used or laboratory results, the diagnoses and species information based on ICD codes cannot be validated. Overestimates may result from using ICD codes reported in discharge records, which includes documentation of all diagnoses and possibly some incorrect diagnoses based on clinical suspicion or rapid diagnostic tests alone without laboratory confirmation or misinterpretation of inappropriate diagnostic tests (e.g., ELISA antibody test), as compared to the more strict definitions used by the CDC surveillance system. Pre-existing conditions or other conditions used to define cases but that are not due to malaria listed on the discharge records may also result in misclassification of severe malaria cases and its distribution of case characteristics. For example, pre-existing epilepsy or other neurologic disease among uncomplicated P. vivax, P. ovale or P. malariae cases may result in misclassification to severe malaria, and lead to the higher than expected numbers of severe malaria cases by those species. The high proportion of clinically imprecise ICD diagnosis codes, such as those corresponding to conditions that can have a wide clinical spectrum (e.g., non-traumatic acute kidney injury), and changes in the criteria for certain conditions over time, may also preclude consistent and accurate identification of relevant clinical manifestations for classifying cases and co-morbidities. Costs of severe malaria hospitalizations may be underestimated, as reported hospital charges do not include physician fees and other ancillaries, and do not account for other aspects of the economic burden such as lost productivity and other social costs. Since the NIS is discharge record and not patient based, hospital transfers and recrudescent malaria cases that result in re-hospitalizations may lead to multiple counting. Imputation of missing variables was not feasible due to the large size of the NIS and limited computing power. Random and intentional (state-level suppression) missing data on race and sensitive conditions (e.g. HIV) [
      • Healthcare Cost and Utilization Project (HCUP)
      Introduction to the HCUP Nationwide Inpatient Sample (NIS) 2011.
      ,
      • Healthcare Cost and Utilization Project (HCUP)
      Introduction to the HCUP Nationwide inpatient sample (NIS) 2012.
      ] may have resulted in some error in the race-specific number estimates, and in the assessment of HIV as a risk factor. Selection bias [
      • Hernán M.A.
      • Robins J.M.
      Causal inference.
      ,
      • Rothman K.J.
      • Last T.L.
      • Greenland S.
      Modern epidemiology.
      ] is possible in the assessment of associations with severe malaria in this study since both severe malaria and the potential predictors of interest may affect the hospitalization rate. In addition, other potential risk factors of interest, including time to presentation, laboratory results, pre-travel consultation, country of origin, travel history, destination and purpose of travel, immigration and immunity status, and vector avoidance and chemoprophylaxis use were not available in NIS data to determine their associations with severe malaria.
      This study provides insight to the burden of severe malaria and the specific malaria complications in the US, and is generalizable to US travelers, foreign visitors, and immigrants. In light of the increases in international travel, primary and secondary prevention measures need to be emphasized in high-risk groups. Clinicians should include malaria in the differential diagnosis when assessing the etiology of unexplained fever or severe infection, and elicit travel history at initial presentation for prompt diagnosis and initiation of appropriate treatment to reduce the risk of disease progression [
      • Newman R.D.
      • Parise M.E.
      • Barber A.M.
      • Steketee R.W.
      Malaria-related deaths among U.S. travelers, 1963–2001.
      ]. The CDC provides information on malaria risk and prevention recommendations by region [

      Centers for Disease Control and Prevention. Malaria. Yellow book 2015. Available at: http://wwwnc.cdc.gov/travel/yellowbook/2016/infectious-diseases-related-to-travel/malaria. [Accessed 15 March 2017].

      ], and has staff that provides free guidance on diagnosis [
      • Ash L.R.
      • Orihel T.C.
      Atlas of human parasitology.
      ] and treatment [
      • Centers for Disease Control and Prevention
      Treatment of malaria.
      ]. Lastly, continued funding for communicable disease control systems and for tropical medicine research efforts will also be important in combating the malaria problem.

      Ethics

      This research did not involve human subjects, and ethical approval was not required or obtained.

      Authorship statement

      All authors made substantial contributions to all of the following: (1) the conception and design of the study, or acquisition of data, or analysis and interpretation of data, (2) drafting the article or revising it critically for important intellectual content, (3) final approval of the version to be submitted.

      Conflicts of interest

      The authors declare that there are no conflicts of interest.

      Funding

      This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

      Provenance and peer review

      Not commissioned; externally peer reviewed.

      Acknowledgements

      We would like to thank Paul Arguin for reviewing this manuscript.

      References

        • World Malaria Report 2016
        World health organization.
        World Health Organization, Geneva, Switzerland2016
        • Murray C.J.L.
        • Ortblad K.F.
        • Guinovart C.
        • Lim S.S.
        • Wolock T.M.
        • Roberts D.A.
        • et al.
        Global, regional, and national incidence and mortality for HIV, tuberculosis, and malaria during 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013.
        Lancet. 2014; 384: 1005-1070
      1. Centers for Disease Control. Malaria. Available at: http://www.cdc.gov/malaria [Accessed 15 April 2017].

        • Cullen K.A.
        • Mace K.E.
        • Arguin P.M.
        Malaria surveillance–United States, 2013.
        MMWR Surveill Summ. 2016; 65: 1-22
        • Checkley A.M.
        • Smith A.
        • Smith V.
        • Blaze M.
        • Bradley D.
        • Chiodini P.L.
        • et al.
        Risk factors for mortality from imported falciparum malaria in the United Kingdom over 20 years: an observational study.
        BMJ. 2012; : 344
        • Hagmann S.H.F.
        • Van P.V.
        • Stauffer W.M.
        • Miller A.O.
        • Connor B.A.
        • Hale D.C.
        • et al.
        Travel-associated disease among US residents visiting US GeoSentinel clinics after return from international travel.
        Fam Pract. 2014; 31: 678-687
        • Odolini S.
        • Gautret P.
        • Parola P.
        Epidemiology of imported malaria in the Mediterranean region.
        Mediterr J Hematol Infect Dis. 2012; 4: e2012031
        • Seringe E.
        • Thellier M.
        • Fontanet A.
        • Legros F.
        • Bouchaud O.
        • Ancelle T.
        • et al.
        Severe imported Plasmodium falciparum malaria, France, 1996–2003.
        Emerg Infect Dis. 2011; 17: 807-813
        • Khuu D.
        • Eberhard M.L.
        • Bristow B.N.
        • Javanbakht M.
        • Ash L.R.
        • Shafir S.C.
        • et al.
        Malaria-related hospitalizations in the United States, 2000-2014.
        Am J Trop Med Hyg. 2017; 97: 213-221
        • Healthcare Cost and Utilization Project (HCUP)
        Introduction to the HCUP Nationwide Inpatient Sample (NIS) 2011.
        Agency for Healthcare Research and Quality, Rockville, MD2013 (Available at:)
        • Healthcare Cost and Utilization Project (HCUP)
        Introduction to the HCUP Nationwide inpatient sample (NIS) 2012.
        Agency for Healthcare Research and Quality, Rockville, MD2014 (Available at:)
      2. Centers for Disease Control and Prevention. Classification of diseases, functioning, and disability. Conversion table. Available at: http://www.cdc.gov/nchs/icd/icd9cm.htm. [Accessed 31 March 2017].

        • HCUP Clinical Classifications Software
        Healthcare cost and utilization Project (HCUP).
        Agency for Healthcare Research and Quality, Rockville, MDApril 2016 (Available at:)
        • 3M APR DRG Classification System
        Methodology overview.
        3M Health Information Systems, 2013 (Available at:)
      3. United States Department of Health and Human Services (US DHHS), Centers for disease control and prevention (CDC), national center for health statistics (NCHS), bridged-race population estimates, United States July 1st resident population by state, county, age, sex, bridged-race, and hispanic origin. Compiled from 1990-1999 bridged-race intercensal population estimates (released by NCHS on 7/26/2004); revised bridged-race 2000-2009 intercensal population estimates (released by NCHS on 10/26/2012); and bridged-race Vintage 2014 (2010-2014) postcensal population estimates (released by NCHS on 6/30/2015). Available on CDC WONDER Online Database. Accessed at http://wonder.cdc.gov/bridged-race-v2014.html on Apr 15, 2016 1:00:23 PM

        • Chalwe V.
        • Van geertruyden Jean-Pierre
        • Mukwamataba D.
        • Menten J.
        • Kamalamba J.
        • Mulenga M.
        • et al.
        Increased risk for severe malaria in HIV-1–infected adults, Zambia.
        Emerg Infect Dis. 2009; 15: 749-755
        • Danquah I.
        • Bedu-Addo G.
        • Mockenhaupt F.P.
        Type 2 diabetes mellitus and increased risk for malaria infection.
        Emerg Infect Dis. 2010; 16: 1601-1604
        • Gallego-Delgado J.
        • Rodriguez A.
        Malaria and hypertension. Another co-evolutionary adaptation?.
        Front Cell Infect Microbiol. 2014; 4: 121
        • Abanyie F.A.
        • Arguin P.M.
        • Gutman J.
        State of malaria diagnostic testing at clinical laboratories in the United States, 2010: a nationwide survey.
        Malar J. 2011; : 340
        • Doyle T.J.
        • Glynn M.K.
        • Groseclose S.L.
        Completeness of notifiable infectious disease reporting in the United States: an analytical literature review.
        Am J Epidemiol. 2002; 155: 866-874
        • Das B.S.
        Renal failure in malaria.
        J Vector Borne Dis. 2008; 45: 83-97
        • Walker D.M.
        • Oghumu S.
        • Gupta G.
        • McGwire B.S.
        • Drew M.E.
        • Satoskar A.R.
        Mechanisms of cellular invasion by intracellular parasites.
        Cell Mol Life Sci. 2014; 71: 1245-1263
        • Hwang J.
        • Cullen K.A.
        • Kachur S.P.
        • Arguin P.M.
        • Baird J.K.
        Severe morbidity and mortality risk from malaria in the United States, 1985–2011.
        Open Forum Infect Dis. 2014; 1 (ofu034)
        • Mendis K.
        • Sina B.J.
        • Marchesini P.
        • Carter R.
        The neglected burden of Plasmodium vivax malaria.
        Am J Trop Med Hyg. 2001; 64: 97-106
        • Trampuz A.
        • Jereb M.
        • Muzlovic I.
        • Prabhu R.M.
        Clinical review: severe malaria.
        Crit Care. 2003; 7: 315-323
        • Kochar D.K.
        • Saxena V.
        • Singh N.
        • Kochar S.K.
        • Kumar S.V.
        • Das A.
        Plasmodium vivax malaria.
        Emerg Infect Dis. 2005; 11: 132-134
        • Ash L.R.
        • Orihel T.C.
        Atlas of human parasitology.
        5th ed. ASCP Press, Chicago2007
        • Mühlberger N.
        • Jelinek T.
        • Behrens R.H.
        • Gjørup I.
        • Coulaud P.
        • Clerinx J.
        • et al.
        Age as a risk factor for severe manifestations and fatal outcome of falciparum malaria in European patients: observations from TropNetEurop and SIMPID Surveillance Data.
        Clin Infect Dis. 2003; 36: 990-995
        • Leder K.
        • Black J.
        • O’Brien D.
        • Greenwood Z.
        • Kain K.C.
        • Schwartz E.
        • et al.
        Malaria in travelers: a review of the GeoSentinel surveillance network.
        Clin Infect Dis. 2004; 39: 1104-1112
        • Coggins C.H.
        • Breyer Lewis J.
        • Caggiula A.W.
        • Castaldo L.S.
        • Klahr S.
        • Wang S.R.
        Differences between women and men with chronic renal disease.
        Nephrol Dial Transplant. 1998; 13: 1430-1437
        • Yu Z.
        • Zhan J.
        • Li C.
        • Zhou H.
        Age and gender analysis of jaundice patients.
        J Biosci Med. 2012; 2: 1-3
        • Käser A.K.
        • Arguin P.M.
        • Chiodini P.L.
        • Smith V.
        • Delmont J.
        • Jimenez B.C.
        • et al.
        Imported malaria in pregnant women: a retrospective pooled analysis.
        Trav Med Infect Dis. 2015; 13: 300-310
        • Hall H.I.
        • An Q.
        • Tang T.
        • Song R.
        • Chen M.
        • Green T.
        • et al.
        Prevalence of diagnosed and undiagnosed HIV infection – United States, 2008–2012.
        MMWR Morb Mortal Wkly Rep. 2015; 64: 657-680
        • Centers for Disease Control and Prevention
        National diabetes statistics report: estimates of diabetes and its burden in the United States, 2014.
        US Department of Health and Human Services, Atlanta, GA2014 (Available at:)
        • Nwankwo T.
        • Yoon S.S.
        • Burt V.
        • Gu Q.
        Hypertension among adults in the United States: national health and nutrition examination survey, 2011–2012. NCHS data brief, no 133.
        National Center for Health Statistics, Hyattsville, MD2013 (Available at:)
        • Hernán M.A.
        • Robins J.M.
        Causal inference.
        Chapman & Hall/CRC, forthcoming, Boca Raton2016
        • Rothman K.J.
        • Last T.L.
        • Greenland S.
        Modern epidemiology.
        3rd ed. Wolters Kluwer Health, Philadelphia2012
        • Newman R.D.
        • Parise M.E.
        • Barber A.M.
        • Steketee R.W.
        Malaria-related deaths among U.S. travelers, 1963–2001.
        Ann Intern Med. 2004; 141: 547-555
      4. Centers for Disease Control and Prevention. Malaria. Yellow book 2015. Available at: http://wwwnc.cdc.gov/travel/yellowbook/2016/infectious-diseases-related-to-travel/malaria. [Accessed 15 March 2017].

        • Centers for Disease Control and Prevention
        Treatment of malaria.
        2013 (Available at:)