Epidemiology and Trends of Necrotizing Enterocolitisin Sweden: 1987–2009

WHAT’S KNOWN ON THIS SUBJECT: The incidence of necrotizingenterocolitis (NEC), a devastating condition in neonates, variesgeographically and with time. Although the most consistent riskfactors are prematurity and low birth weight, it has not beenconvincingly shown to increase in the postsurfactant era.

WHAT THIS STUDY ADDS: The incidence of NEC, especially amongthe highly premature but also in more mature groups, hasincreased in recent decades, concurrent with dramaticallyimproved early infant survival. Seasonal variation of NEC suggestsenvironmental etiological factors.

abstractOBJECTIVE: To investigate temporal, seasonal, and geographic varia-tions in the incidence of necrotizing enterocolitis (NEC) and its re-lation to early infant survival in the Swedish population and insubgroups based on gestational age, birth weight, and gender.

METHODS: In the Swedish birth cohort of 1987 through 2009 all chil-dren with a diagnosis of NEC were identified in the National PatientRegister, the Swedish Medical Birth Register, and the National Causeof Death Register. NEC incidence, early mortality, and seasonality wereanalyzed with descriptive statistics, Poisson regression, and auto re-gression.

RESULTS: The overall incidence of NEC was 3.4 in 10 000 live births,higher in boys than in girls (incidence rate ratio 1.22, 95% confidenceinterval 1.06–1.40, P = .005), with a peak in November and a trough inMay, and increased with an average of ∼5% a year during the studyperiod. In most subgroups, except the most immature, an initial de-crease was followed by a steady increase. Seven-day mortalitydecreased strongly in all subgroups over the entire study period(annual incidence rate ratio 0.96, 95% confidence interval 0.95–0.96,P, .001). This was especially marked in the most premature and lowbirth weight infants.

CONCLUSIONS: After an initial decrease, the incidence of NEC has in-creased in Sweden during the last decades. An association with theconcurrent dramatically improved early survival seems likely. Pediat-rics 2013;132:1–9

AUTHORS: Margareta Ahle, MD,a Peder Drott, MD, PhD,b

and Roland E. Andersson, MD, PhDc

aDivision of Radiology, Department of Medical and HealthSciences, Faculty of Health Sciences, Linköping University;Department of Radiology in Linköping, County Council ofÖstergötland, Linköping, Sweden; bDivision of Surgery,Department of Clinical and Experimental Medicine, Faculty ofHealth Sciences, Linköping University, Linköping; Department ofSurgery, Section of Pediatric Surgery, County Council ofÖstergötland, Linköping, Sweden; cDivision of Surgery,Department of Clinical and Experimental Medicine, Faculty ofHealth Sciences, Linköping University, Linköping, Department ofSurgery, Ryhov County Hospital, Jönköping, Sweden

KEY WORDSnecrotizing enterocolitis, premature infants, perinatal mortality,perinatal care, epidemiology, trends, seasonal variation

ABBREVIATIONSCI—confidence intervalICD-9—International Classification of Diseases, Ninth RevisionICD-10—International Classification of Diseases, 10th RevisionIRR—incidence rate ratioNCD—National Cause of Death RegisterNEC—necrotizing enterocolitisNPR—National Patient RegisterSMB—Swedish Medical Birth Register

Dr Ahle initiated the study, participated in the design of thestudy and the request of the database extract, processed andanalyzed the data, drafted the initial manuscript, and approvedthe final manuscript as submitted. Dr Drott suggested the fieldof study, took part in the discussion of study design, definition ofterms, and interpretation of results, reviewed and revised themanuscript; and approved the final manuscript as submitted. DrAndersson suggested the method, participated in the design ofthe study and the request of the database extract, carried outsome of the analyses, reviewed and revised the manuscript, andapproved the final manuscript as submitted.

www.pediatrics.org/cgi/doi/10.1542/peds.2012-3847

doi:10.1542/peds.2012-3847

Accepted for publication Apr 25, 2013

Address correspondence to Margareta Ahle, MD, Department ofRadiology, University Hospital, 581 85 Linköping, Sweden. E-mail:Margareta.Ahle@lio.se

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright © 2013 by the American Academy of Pediatrics

FINANCIAL DISCLOSURE: The authors have indicated they haveno financial relationships relevant to this article to disclose.

FUNDING: Supported by the County Council of Östergötland,Futurum, the Academy of Health Care, Jönköping County Council,Jönköping, Sweden, and the Medical Research Council ofSoutheast Sweden.

PEDIATRICS Volume 132, Number 2, August 2013 1

ARTICLE

by guest on June 23, 2018www.aappublications.org/newsDownloaded from

Survival of preterm infants has increaseddramatically over thepastdecadesdue toimproved perinatal and neonatal care.There are concerns that this may leadto increased incidence rates of morbid-ities specific to the premature, for ex-ample, necrotizing enterocolitis (NEC).1–4

Reports of trends in the incidence ofNEC are, however, contradictory, showingboth decreasing and increasing rates(Table 1).1–7

NEC incidence is known to vary geo-graphically with frequencies thought tobe especially low in Japan, Switzerland,and some Scandinavian countries.8–10

Most studies are from the United Statesand report incidence rates between 0.7and 1 in 1000 live birthswith significantlyhigher incidence rates in infants of Afri-can American descent (1.6–2.2 per 1000live births) than whites (0.5–0.6 per 1000live births).1,11–13

In the United Kingdom and Ireland, 0.3cases of proven NEC in 1000 live birthswere estimated (0.5 when including sus-pected cases).14 Among infantswith birthweights of 1500 g or less, NEC wasreported in 6% in the Netherlands in1983, 2% in Italy from 1994 through 2005,and 7% in Spain from 2002 through2005.15–17

To our knowledge, there is no truly pop-ulation based study from Europe, andmost studies are based only on selectedinfants with low birth weights and pre-maturegestationalages. Inthedatabasesheld by the Swedish National Board ofHealth andWelfare18–22 datawith respectto NEC are available since 1987, whena specific code for NEC was introduced

with the ninth edition of the World HealthOrganization’s International Classifica-tion of Diseases (ICD-9). This allows fora survey of a long period of time in whichmajor changes in neonatal care havetaken place.

The current study is a population-basedstudy of the incidence of NEC in Swedenfrom 1987 through 2009. We describetemporal, seasonal, and geographic var-iations, as well as associations withgestational age, birthweight, and gender.To address the question of potentiallyincreasing NEC incidence in relation toimprovement in ante- and perinatal care,we also analyze the development of earlymortality.

METHODS

Data were obtained from the followingregisters held by the Swedish NationalBoard of Health andWelfare: the NationalPatient Register (NPR),18,21 the SwedishMedical Birth Register (SMB),19,22 andthe National Cause of Death Register(NCD).20 All children born in 1987through 2009 with a diagnosis of NEC(ICD-9 or the 10th revision of the ICD[ICD-10] code 777F or P77, respectively),in the NPR, NCD, or SMB were identified.Gestational age, weight at birth, and otherperinatal data were obtained by linkagewith the SMB based on the children’sunique national Personal Identity Numberassigned to .99.9% of all Swedish resi-dents at birth or immigration.23

Because of a few weeks’ delay in re-ceiving the identity number after birth,this number was missing for someregistrations in the NPR and NCD,

which impeded linkage with the SMB.Some of these discharges representa chain of consecutive admissions ofthe same child because of referralsbetween hospitals of different levels ofcare and repeated admissions becauseof complications. On the basis of reg-istry data on gender, date of birth,dates of discharge and admission,place of residency, diagnoses otherthan NEC, as well as knowledge of thereferral system between hospitals, wewere able to trace and link most of theadmissions to unique individuals. Spe-cial care was taken with multiplebirths as marked in the SMB variableand/or detected by ICD codes in theNPR. The study was approved by theRegional Ethical Review Board ofLinköping.

The background population consists ofall children born in Sweden from 1987through 2009. A deidentified extract wasobtained from the SMB Register con-taining demographic data and perinatalinformation including date of birth,gender,gestationalage,andbirthweight.

Subgroups were defined by gender,gestational age, and birth weight. Limitswere chosen to fit definitions of ICD-10.The number of children belonging to thelowest weight classes specified by spe-cialist codes (,500 g and 500–749 g),however, were too small to allow formeaningful statistical analysis and weretherefore brought together in the sub-groupwith a birthweight,750 g. Amongchildren with a birth weight.4500 g ora gestational age .42 weeks, NEC is ex-tremely rare, which justified merging

TABLE 1 Development of NEC Incidence Over Time According to Earlier Studies

Country Population Study Periods NEC Frequency during study periods Reference

United States Nonselected 1986 1992 Signs of increasing rates 2,3Australia 24–28 wk gestation 1986–1987; 1992–1993 1998–1999 9%–10% 5% 4Sweden Birth wt ,1000 g 1990–1992 2% 5United States Nonselected 1991–1994 1995–1998 Relative risk 1.6 for the second period compared with the first 1Sweden ,32 wk gestation 1995–1999 2000–2004 1.2% 1.2% 6Sweden ,27 wk gestation 2004–2007 6% 7

2 AHLE et al by guest on June 23, 2018www.aappublications.org/newsDownloaded from

these groups with the groups of full termand normal weight births, respectively.

The study period was divided into 4subperiods of 6, 6, 6, and 5 years, re-spectively. The incidence of NEC per10 000 live births, including trends overtime as well as geographic and sea-sonal variation, was analyzed for theentire population and for subgroups.Trends of early mortality, defined asdeathwithin 7 days of birth per 1000 livebirths, were also analyzed in the pop-ulation as awhole and in the subgroups.

Statistical Methods

Data were analyzed with descriptivestatistics. To obtain robust estimatesand confidence intervals (CIs) of thecount data, Poisson regression wasused for trends and covariations andresults presented as incidence rateratio (IRR).24 To describe the strength oftrends over time, the term “annual IRR”is used as ameasure of average annualchange in incidence rates. Seasonalvariation was analyzed by investigatingfor the presence of harmonic cyclicitywith auto correlation. The eventualpeak and trough were identified graphi-cally by a seasonal subseries plot. Thecomplete monthly time series was usedfor these analyses.

Missing Data

Individuals with NEC for whom the peri-natal data of the SMB could not betraced because of incomplete identityinformation in the NPR and NPD, as de-scribed earlier, were excluded. To eval-uate the influence of the these cases onthe epidemiologic results, the distribu-tion of the diagnosis “highly premature”or “very low birth weight” (765A ac-cording to ICD-9 or P070 or P072 ac-cording to ICD-10), “other” premature orlow birthweight (ICD-9 765B; ICD-10 P071or P073), and “small for gestational age”(ICD-9 764A, 764B, 764D, 764E or 7643;ICD-10P05.0 or P051) aswell as in-hospitaldeaths was compared for the children

with and without traceable perinataldata in the SMB.

RESULTS

Patients

During the studied period, 2 381 318 livebirths were registered in SMB, 1 223 501(51%) boys and 1 157 387 (49%) girls. In430 infants, information on gender wasmissing.

Nine hundred twenty cases of NEC, 521(57%) boys and 399 (43%) girls, werefoundin theNPR,NCD,and/orSMB.For112cases, 66 (59%) boys and 46 (41%) girls,theperinatal information in theSMBcouldnot be traced. A diagnosis of prematuritywas less frequent among these casescompared with the 808 cases for whomthe perinatal information could be traced(64%vs80%,P, .001), but theproportiondiagnosed as highly premature wassimilar (38% vs 42%, P = .36). The groupwith missing perinatal information makeup 11% of all NEC cases diagnosed as“highly premature” (765A, P070, or P072)and 9% of those diagnosed as “otherpremature” (765B, P071, or P073; P = .31).

Incidence of NEC

The overall incidence of NEC during thestudy period was 3.4 in 10 000 live births(3.8 in 10 000 including the cases withincomplete identity information; Table 2).There were strong associations withprematurity and low birth weights, but18% of all the NEC patients were full termand 19% had a birth weight of$2500 g.

The incidence was higher in boys than ingirls (3.7 vs 3.0/10 000, P = .02 or IRR 1.22,95%CI 1.06–1.40, P= .005). This differenceremained after adjustment for birthweight, gestational age, and birth year(IRR 1.25, 95% CI 1.08–1.43, P = .002). Itwas especially marked in children withbirth weights 1000 to 1499 g (183.2 vs119.3/10 000, P = .006) and 1500 to 2499 g(22.7 vs 14.7/10 000,P= .006) but not seenamong the children with gestational age,28 weeks or birth weight,1000 g.

Except forahigher frequencyofNEC inthecounty of Stockholm, the capital of Swe-den and its largest city, there were nodifferences in incidence rates accordingto region or population density. The in-creased incidence rates of StockholmCounty compared with the rest of thecountry were seen in all subgroups withminor fluctuations over time. However,the incidence of NEC in children that diedwithin28dayswasthesameinStockholmas in the rest of the country.

A certain seasonal variation in NEC in-cidence rates was seen with a peak inNovember and a trough in May, whichpersisted after adjustment for pre-maturity and was confirmed by auto re-gression.

Trends

The incidence curve of NEC showed aj-shaped pattern with an initial de-creasebetween thefirst and the secondsubperiod (1987–1992 and 1993–1998),followed by a steady increase (Table 3,Figs 1 and 2). This pattern was seen forall subgroups of gestational age andbirth weight, except in gestational ages,28 full weeks and birth weights,1000 g, in which NEC incidence ratesincreased over the entire period. Theproportion of term infants among NECcases decreased from 30% in the firstsubperiod to 14% in the last (P, .001).

From the first to the last subperiod, theproportions of infants with gestationalages ,28 weeks and birth weights,750 g increased in the backgroundpopulation from 24.4 to 30.4 per 10 000live births (IRR 1.25, 95%CI 1.16–1.34, P,.001) and from 9.0 to 14.4 per 10 000 livebirths (IRR 1.61, 95% CI 1.45–1.79, P ,.001), respectively. For the other sub-groups, there were only small changesover time.

During the study period, 7-day mortalitydecreased strongly (annual IRR 0.96, 95%CI 0.95–0.96, P, .001; Table 4, Figs 3 and4). The trendwas similar in all subgroupsof birth weight, and gestational age, with

ARTICLE

PEDIATRICS Volume 132, Number 2, August 2013 3 by guest on June 23, 2018www.aappublications.org/newsDownloaded from

the strongest decrease in the groupswith birthweights of 750 to 999 g (annualIRR 0.93, 95% CI 0.92–0.95, P, .001) andgestational ages 28 to 31 weeks (annualIRR = 0.94, 95% CI 0.92–0.95, P , .001),corresponding to reductions of 66% and63%respectively from the first subperiodto the last.

DISCUSSION

This study shows a considerable in-crease in NEC incidence together with

strongly decreasing infant mortality inSweden during 1987 through 2009,a period of major changes in perinatalcare. The first case reports of NEC werepublished in the first half of the 19thcentury, but itwaswith thedevelopmentof neonatal intensive care in the 1960sthat the condition became more com-mon and was described as a clinical,radiographic, and pathologic entity.8 InSweden, the routine use of antenatalsteroids and surfactant was gradually

introduced during the 1990s. Since2001, national guidelines for manage-ment of extremely preterm births,apart from centralization of perinatalservices, encourage individualized eval-uation rather than a fixed gestationalage limit for intervention, and from aninternational perspective, perinatal andneonatal care in Sweden is compara-tively proactive.7 The national ExtremelyPreterm Infants Study reports 70%1-year survival among infants with,27 full weeks’ gestation during 2004through 2007,7,25 compared with a sur-vival rate of 48% in a historical cohort ofextremely low birth weight infants bornin Sweden in April 1990 through March1992.5

Our results, in contrast to others,4 revealthat concerns about a possible rise inincidence rates of NEC1–4 are justified, butthe mechanisms behind this observationseem to be complex, with improved sur-vival of the most premature as 1 of manycontributing factors. Whereas early mor-tality has been steadily decreasing in

TABLE 2 Distribution of the Background Population of All Live Births in Sweden and the 808 NEC Patients With Available Perinatal Data

Characteristic Background Population, n n Incidence Per 10 000 Live Births IRR (95% CI)

All 2 381 318 808 3.4Gestational ageFull term 2 232 308 145 0.6 1 Reference32–36 wk 124 307 138 11.1 17.09 (13.54–21.58)28–31 wk 14 822 220 148.4 228.51 (185.29–281.81),28 wk 6595 304 461.0 709.65 (582.29–864.87)

Birth wt, g$2500 2 271 751 157 0.7 1 Reference1500–2499 85 419 158 18.5 26.76 (21.46–33.38)1000–1499 11 047 168 152.1 220.05 (177.03–273.53)750–999 3793 158 416.6 602.75 (483.30–751.72),750 2768 147 531.1 768.44 (613.65–962.29)

SexGirls 1 157 387 353 3.0 1 ReferenceBoys 1 223 501 455 3.7 1.22 (1.06–1.40)

Time period1987–1992 697 229 179 2.6 1 Reference1993–1998 597 876 131 2.2 0.85 (0.68–1.07)1999–2004 558 948 198 3.5 1.38 (1.13–1.69)2005–2009 527 265 300 5.7 2.22 (1.84–2.67)

Place of birthSweden except Stockholm County 1 847 996 506 2.7 1 ReferenceStockholm County 533 322 302 5.6 2.07 (1.79–2.38)

SeasonSummer (April–September) 1250215 365 2.9 1 ReferenceWinter (October–March) 1139400 443 3.9 1.33 (1.16–1.53)

In a few registrations, some information is missing, which explains that numbers do not always add up. Incidence of NEC per 10 000 live births and IRRs are also given.

TABLE 3 Incidence of NEC per 10 000 Live Births for the 4 Subperiods Studied and the TrendExpressed as Annual IRR

Characteristic 1987–1992 1993–1998 1999–2004 2005–2009 Annual IRR (95% CI)

All 2.6 2.2 3.5 5.7 1.05 (1.04–1.06)Gestational ageFull term 0.8 0.4 0.5 0.8 0.99 (0.97–1.02)32–36 wk 8.9 8.4 10.8 17.7 1.04 (1.01–1.06)28–31 wk 124.5 100.2 157.6 224.1 1.04 (1.02–1.06),28 wk 230.2 275.7 492.7 856.3 1.08 (1.06–1.10)

Birth wt, g$2500 0.8 0.5 0.6 0.9 1.00 (0.97–1.02)1500–2499 17.4 11.7 16.8 29.9 1.03 (1.01–1.05)1000–1499 104.3 92.9 155.5 289.0 1.08 (1.05–1.10)750–999 199.4 295.4 475.7 763.6 1.07 (1.05–1.10),750 272.0 320.5 553.4 894.7 1.08 (1.05–1.11)

4 AHLE et al by guest on June 23, 2018www.aappublications.org/newsDownloaded from

all subgroups for the entire study pe-riod, the NEC incidence curve is j-shapedin all groups but those with gestationalages,28 full weeks and birth weights

,1000 g, where an increasing trend isseen throughout the period. Holmansuggested that improved perinatal careinitially helped reduce NEC incidence

before the continuously increasing sur-vival of the most vulnerable individualscaused a subsequent rise.2,3 Preventivemeasures that may have helped reduce

FIGURE 1NEC incidence in subgroups according to gestational age. Note that the scale of the y-axis differs between the subgroups.

FIGURE 2NEC incidence in subgroups according to birth weight. Note that the scale of the y-axis differs between the subgroups.

ARTICLE

PEDIATRICS Volume 132, Number 2, August 2013 5 by guest on June 23, 2018www.aappublications.org/newsDownloaded from

NEC incidence when introduced duringthe first subperiod include scheduledfeedings and the use of humanmilk.26–28

There are also data to suggest thatsurfactant, at least in repeated doses,may prevent NEC.29

It is generally accepted that 90% of allpatients with NEC are premature.13 Theproportion of term infants reportedfrom various studies, however, variesfrom 7% to 25%.1,14,30–34 Possiblyreflecting a relatively low frequency ofprematurity in Sweden, 30% of NEC

cases were full term during our firstsubperiod. The decrease to 14% duringthe last period can be explained bya combination of an increasing pro-portion of premature infants anda more rapidly rising incidence rate inthe extremely premature and low birthweight infants. The latter supports therole of early survival in these groups.Nevertheless, NEC incidence tends torise in all groups. Starting from thesubperiod with the lowest incidencerate (1993–1998), the annual IRR is 1.05

(95% CI 1.00–1.09, P = .04) in full-terminfants. Only in birth weights $2500 gdoes the trend not reach statisticalsignificance. It could be that, also in themature groups, an improved initialsurvival of individuals with conditionsmaking them susceptible to NEC playsa role in the rising incidence rate, butthe impact of specific risk factorsneeds additional investigation and willbe addressed in a forthcoming case-control study of the same population.

The incidence rate in the population asa whole is lower than that reportedfrom the United States,1,11–13 but, es-pecially in the last subperiod, similarto United Kingdom and Ireland14 andthe Caucasian population in the UnitedStates.1,13 The overall incidence amonginfants of our population with birthweights ,1500 g (268.6/10 000) ishigher than in the Italian study,15 es-pecially during the last subperiod(507.3/10 000) but lower than in theDutch and Spanish populations.16,17

Why NEC incidence varies betweencountries is not fully understood, but

TABLE 4 Seven-Day Mortality per 1000 Live Births for the 4 Subperiods Studied, and the TrendExpressed as Annual IRR

Characteristic 1987–1992 1993–1998 1999–2004 2005–2009 Annual IRR (95% CI)

All 2.82 2.16 1.74 1.27 0.96 (0.95–0.96)Gestational ageFull term 0.96 0.72 0.61 0.40 0.96 (0.95–0.96)32–36 wk 11.34 8.81 7.1 5.08 0.96 (0.95–0.97)28–31 wk 70.47 47.41 32.91 25.78 0.94 (0.93–0.95),28 wk 348.29 256.27 180.65 153.75 0.95 (0.94–0.96)

Birth wt, g$2500 g 0.90 7.16 0.57 0.39 0.95 (0.95–0.96)1500–2499 g 15.42 10.37 8.75 6.24 0.95 (0.94–0.96)1000–1499 g 87.58 58.21 42.38 32.36 0.94 (0.93–0.96)750–999 g 229.82 142.41 75.49 78.79 0.93 (0.92–0.95),750 g 502.40 395.83 275.36 231.58 0.96(0.95–0.97)

FIGURE 3Early mortality in subgroups according to gestational age. Note that the scale of the y-axis differs between the subgroups.

6 AHLE et al by guest on June 23, 2018www.aappublications.org/newsDownloaded from

differences in ethnicity may play a rolealongside great differences in theproportion of premature births.35

Seasonal variation in NEC incidence hasseldom been studied. Snyder found aperiodicityof6monthswithpeaks inJuneand December.36 In contrast, our datasuggest a periodicity of 12 months witha peak in November and a trough in May.This may indicate that there are envi-ronmental factors with seasonal varia-tion influencing the incidence of NEC butthat these factors do not occur in thesame patterns in different populations.

Our study is comparatively large, trulypopulation based, and stretches overa considerable period of time, includingtheyearsbeforethegradual introductionof antenatal steroids and surfactant inroutine perinatal care during the 1990s.7

The registers, held by the Swedish Na-tional Board of Health and Welfare,provide high-quality information on allhospital discharges in Sweden, whichentail good conditions for comprehen-sive population-based epidemiologic

studies.21,22 However, certain difficulties,some of which are especially obvious inthe newborn, arise from the use ofpassive rather than active case ascer-tainment methodology.

During the study period, newborn in-fants received their identification num-bers with some weeks delay, and thereare 112 cases of NEC, representing 12%of all cases, for whom the perinatal datacould not be obtained from the SMB. Adischarge diagnosis indicating pre-maturity was less common among thecases with missing perinatal informa-tion, but there was no difference in theproportion of highly premature or in-hospital deaths. Therefore, it seemsreasonable to assume that these miss-ing cases do not induce any substantialselection bias in the analyses based onperinatal data. Missing identificationnumbers are less frequent during thelast subperiod than before, especially inboys. The overall positive incidencetrendfor thewholepopulation,aswellasforboysandgirlsseparately, is,however,

not significantly affected if the caseswith missing perinatal information areincluded or not.

The limitations of hospital dischargedata, registered as part of an admin-istrative routine rather than accordingto a research protocol with strict di-agnostic criteria to fit the issue ofa specific study, must also be taken intoconsideration when interpreting themajor results of the study. The regionaldifferences, with similar incidencethroughout the country among thechildren that died within 28 days buthigher incidence rate in Stockholmamong the survivors, indicate that lesssevere cases have been diagnosed andmay illustrate a possible influence fromdifferences in registration praxis. Incases with long hospitalizations andmany conditions, a minor episode ofNEC, healed long before discharge, maybe forgotten when diagnoses arereported. Conversely, because no in-formation is available on the criteriaused for diagnosing NEC, suspected

FIGURE 4Early mortality in subgroups according to birth weight. Note that the scale of the y-axis differs between the subgroups.

ARTICLE

PEDIATRICS Volume 132, Number 2, August 2013 7 by guest on June 23, 2018www.aappublications.org/newsDownloaded from

cases will have been included alongwith certain ones. On the other hand,criteria requiring a positive radiographmay miss a substantial proportion ofreal cases.37 The recent discussion ondifferential diagnosis in NEC could in-fluence the incidence rates over timebecause some cases might get otherdiagnoses toward the end of the studyperiod.38 This, however, should causedecreasing incidence rates rather thanincreasing. With an augmented aware-ness of NEC, the readiness to suspectthe diagnosismay increase over time. Inthat case, however, less serious caseswould be included and NEC incidencewould rise faster in those who survived

28 days than in those who did not, butthe trend did not differ significantlybetween the groups: annual IRR, ad-justed for birth weight and gestationalage, 1.04 (95% CI 1.03–1.05, P, .01) forsurvivors and 1.06 (95% CI 1.04–1.09,P, .01) for nonsurvivors, and a positivetrend in the proportion of live born dy-ing from NEC as main or contributingcause parallels the incidence trend:annual IRR 1.05 (1.02–1.08, P = 0001).

CONCLUSIONS

We have shown an important increasein the incidence rate of NEC in Swedenfrom 1987 through 2009 alongside a

dramatically improved survival, espe-cially for the premature and low birthweight infants. Because early survivalof the most vulnerable infants con-tinues to improve, the incidence ratesof NEC most probably will continueto rise. Differentiated epidemiologicknowledge and improved understand-ing of the pathogenesis and patho-physiology of NEC are needed toachieve effective prevention strategiesand early detection of individuals atrisk.

ACKNOWLEDGMENTWe thank Almut Herzog, PhD, for helpwith the database work.

REFERENCES

1. Llanos AR, Moss ME, Pinzòn MC, Dye T,Sinkin RA, Kendig JW. Epidemiology ofneonatal necrotising enterocolitis: a pop-ulation-based study. Paediatr Perinat Epi-demiol. 2002;16(4):342–349

2. Holman RC, Stoll BJ, Clarke MJ, Glass RI.The epidemiology of necrotizing enteroco-litis infant mortality in the United States.Am J Public Health. 1997;87(12):2026–2031

3. Holman RC, Stehr-Green JK, Zelasky MT.Necrotizing enterocolitis mortality in theUnited States, 1979–85. Am J Public Health.1989;79(8):987–989

4. Luig M, Lui K; NSW and ACT NICUS Group.Epidemiology of necrotizing enterocolitis—Part I: changing regional trends in ex-tremely preterm infants over 14 years.J Paediatr Child Health. 2005;41(4):169–173

5. Finnström O, Olausson PO, Sedin G, et al.The Swedish national prospective study onextremely low birthweight (ELBW) infants.Incidence, mortality, morbidity and survivalin relation to level of care. Acta Paediatr.1997;86(5):503–511

6. Lundqvist P, Källén K, Hallström I, Westas LH.Trends in outcomes for very preterm infantsin the southern region of Sweden over a 10-year period. Acta Paediatr. 2009;98(4):648–653

7. EXPRESS Group. Incidence of and risk fac-tors for neonatal morbidity after activeperinatal care: Extremely Preterm InfantsStudy in Sweden (EXPRESS). Acta Paediatr.2010;99(7):978–992

8. Kosloske AM. Epidemiology of necrotizingenterocolitis. Acta Paediatr Suppl. 1994;396(suppl 396):2–7

9. Sankaran K, Puckett B, Lee DSC, et al; Cana-dian Neonatal Network. Variations in incidenceof necrotizing enterocolitis in Canadian neo-natal intensive care units. J Pediatr Gastro-enterol Nutr. 2004;39(4):366–372

10. Kawase Y, Ishii T, Arai H, Uga N. Gastroin-testinal perforation in very low-birthweightinfants. Pediatr Int. 2006;48(6):599–603

11. Guner YS, Friedlich P, Wee CP, Dorey F,Camerini V, Upperman JS. State-basedanalysis of necrotizing enterocolitis out-comes. J Surg Res. 2009;157(1):21–29

12. Holman RC, Stoll BJ, Curns AT, Yorita KL,Steiner CA, Schonberger LB. Necrotisingenterocolitis hospitalisations among neo-nates in the United States. Paediatr PerinatEpidemiol. 2006;20(6):498–506

13. Wilson R, Kanto WP Jr, McCarthy BJ, et al.Epidemiologic characteristics of necrotiz-ing enterocolitis: a population-based study.Am J Epidemiol. 1981;114(6):880–887

14. Palmer SR, Biffin A, Gamsu HR. Outcome ofneonatal necrotising enterocolitis: resultsof the BAPM/CDSC surveillance study,1981–84. Arch Dis Child. 1989;64:388–394

15. Bertino E, Giuliani F, Prandi G, Coscia A,Martano C, Fabris C. Necrotizing entero-colitis: risk factor analysis and role ofgastric residuals in very low birth weightinfants. J Pediatr Gastroenterol Nutr.2009;48(4):437–442

16. Walther FJ, Verloove-Vanhorick SP, Brand R,Ruys JH. A prospective survey of necrotis-ing enterocolitis in very low birthweightinfants. Paediatr Perinat Epidemiol. 1989;3(1):53–61

17. Moro M, Pérez-Rodriguez J, Figueras-Aloy J,et al. Predischarge morbidities in ex-tremely and very low-birth weight infantsin Spanish neonatal units. Am J Perinatol.2009;26(5):335–343

18. National Patient Register (in English).Available at: http://www.socialstyrelsen.se/register/halsodataregister/patientregistret/inenglish. Accessed December 8, 2012

19. Swedish Medical Birth Register (in English).Available at: http://www.socialstyrelsen.se/register/halsodataregister/medicinskafodelseregistret/inenglish. Accessed December8, 2012

20. Dödsorsaksregistret (not available in English).Available at: http://www.socialstyrelsen.se/register/dodsorsaksregistret. AccessedDecember 8, 2012

21. Ludvigsson JF, Andersson E, Ekbom A, et al.External review and validation of theSwedish national inpatient register. BMCPublic Health. 2011;11:450

22. The Swedish National Board of Health andWelfare. Available at: http://www.socialstyrelsen.se/Lists/Artikelkatalog/Attachments/10655/2003-112-3_20031123.pdf. Accessed December8, 2012

23. Ludvigsson JF, Otterblad-Olausson P, PetterssonBU, Ekbom A. The Swedish personal identitynumber: possibilities and pitfalls in healthcareand medical research. Eur J Epidemiol. 2009;24(11):659–667

24. Coxe S, West SG, Aiken LS. The analysis ofcount data: a gentle introduction to Poissonregression and its alternatives. J Pers As-sess. 2009;91(2):121–136

8 AHLE et al by guest on June 23, 2018www.aappublications.org/newsDownloaded from

25. Fellman V, Hellström-Westas L, Norman M,et al; EXPRESS Group. One-year survival of ex-tremely preterm infants after active perinatalcare in Sweden. JAMA. 2009;301(21):2225–2233

26. Wiedmeier SE, Henry E, Baer VL, Stoddard,Lambert DK, Christensen RD. Center dif-ferences in NEC within one health-caresystem may depend on feeding protocol.Am J Perinatol. 2008;23(4):278–285

27. Kosloske AM. Breast milk decreases therisk of neonatal necrotizing enterocolitis.Adv Nutr Res. 2001;10:123–137

28. Sisk PM, Lovelady CA, Dillard RG, Gruber KJ,O’Shea TM. Early human milk feeding isassociated with a lower risk of necrotizingenterocolitis in very low birth weightinfants. J Perinatol. 2007;27(7):428–433

29. Soll R, Özek E. Multiple versus single dosesof exogenous surfactant for the prevention

or treatment of neonatal respiratory dis-tress syndrome. Cochrane Database SystRev. 2009;(1)CD000141

30. Thilo EH, Lazarte RA, Hernandez JA. Necro-tizing enterocolitis in the first 24 hours oflife. Pediatrics. 1984;73(4):476–480

31. Andrews DA, Sawin RS, Ledbetter DJ, SchallerRT, Hatch EI. Necrotizing enterocolitis in termneonates. Am J Surg. 1990;159(5):507–509

32. Ostlie DJ, Spilde TL, St Peter SD, et al. Nec-rotizing enterocolitis in full-term infants. JPediatr Surg. 2003;38(7):1039–1042

33. Beeby PJ, Jeffery H. Risk factors fornecrotising enterocolitis: the influence ofgestational age. Arch Dis Child. 1992;67(4spec no):432–435

34. Henry MCW, Moss RL. Neonatal necrotizingenterocolitis. Semin Pediatr Surg. 2008;17(2):98–109

35. MacDorman MF, Mathews TJ. Behind in-ternational rankings of infant mortality: howthe United States compares with Europe. IntJ Health Serv. 2010;40(4):577–588

36. Snyder CL, Hall M, Sharma V, St Peter SD.Seasonal variation in the incidence ofnecrotizing enterocolitis. Pediatr Surg Int.2010;26(9):895–898

37. Tam AL, Camberos A, Applebaum H. Sur-gical decision making in necrotizingenterocolitis and focal intestinal perfo-ration: predictive value of radiologicfindings. J Pediatr Surg. 2002;37(12):1688–1691

38. Gordon PV, Swanson JR, Attridge JT, ClarkR. Emerging trends in acquired neonatalintestinal disease: is it time to abandonBell’s criteria? J Perinatol. 2007;27(11):661–671

ARTICLE

PEDIATRICS Volume 132, Number 2, August 2013 9 by guest on June 23, 2018www.aappublications.org/newsDownloaded from

originally published online July 1, 2013; Pediatrics Margareta Ahle, Peder Drott and Roland E. Andersson

2009−Epidemiology and Trends of Necrotizing Enterocolitis in Sweden: 1987

ServicesUpdated Information &

012-3847http://pediatrics.aappublications.org/content/early/2013/06/26/peds.2including high resolution figures, can be found at:

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlin its entirety can be found online at: Information about reproducing this article in parts (figures, tables) or

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on June 23, 2018www.aappublications.org/newsDownloaded from

originally published online July 1, 2013; Pediatrics Margareta Ahle, Peder Drott and Roland E. Andersson

2009−Epidemiology and Trends of Necrotizing Enterocolitis in Sweden: 1987

http://pediatrics.aappublications.org/content/early/2013/06/26/peds.2012-3847located on the World Wide Web at:

The online version of this article, along with updated information and services, is

ISSN: 1073-0397. 60007. Copyright © 2013 by the American Academy of Pediatrics. All rights reserved. Print the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois,has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

by guest on June 23, 2018www.aappublications.org/newsDownloaded from