Use of non-steroidal anti-inflammatory drugs and proton pump inhibitors in correlation with incidence, recurrence and death of peptic ulcer bleeding: an ecological study.

0.9 MB·English

by  Lu, Yunxia

Open Access Research
fl
Use of non-steroidal anti-in ammatory
drugs and proton pump inhibitors in
correlation with incidence, recurrence
and death of peptic ulcer bleeding: an
ecological study
Yunxia Lu,1 Emma Sverdén,2 Rickard Ljung,1,3 Claes Söderlund,2
Jesper Lagergren1,4
Tocite:LuY,SverdénE, ABSTRACT
ARTICLE SUMMARY
LjungR,etal.Useofnon- Background:Non-steroidalanti-inflammatorydrugs
steroidalanti-inflammatory
(NSAIDs)andprotonpumpinhibitors(PPIs)are Article focus
drugsandprotonpump
inhibitorsincorrelationwith regardedastwotypesofdrugsthatrespectively Non-steroidal anti-inflammatory drugs (NSAIDs)
incidence,recurrenceand increaseanddecreasetheriskofpepticulcerbleeding. andproton pumpinhibitors (PPIs) are two types of
deathofpepticulcer However,theirrelationtooccurrence,recurrenceand drugs that are closely associated with opposite risk
bleeding:anecologicalstudy. deathofbleedinginthepopulationlevelisnotclear. of peptic ulcer bleeding. Their relation to occur-
BMJOpen2013;3:e002056. Studyobjective:Toclarifyrecentcalendar-time rence, recurrence and death of bleeding in the
doi:10.1136/bmjopen-2012- correlationsbetweensalesofNSAIDsandPPIsandthe populationlevelisnotclear.
002056 occurrenceofpepticulcerbleeding,re-bleedingand
Key messages
▸ Prepublicationhistoryfor death.
thispaperareavailable Design:Ecologicalstudy. ▪ Thesales of NSAIDsand PPIs (inversely) correl-
ate with recurrence and death after peptic ulcer
online.Toviewthesefiles Results:Thetimetrendofpepticulcerbleedingdid
pleasevisitthejournalonline notcorrelatewithPPIsalesbutdidcorrelatewith bleedingonthepopulationlevel.
(http://dx.doi.org/10.1136/ NSAIDsinmem(R =0.6571,P =0.05).Salesof ▪ Women appeared to have a higher fatality of
bmjopen-2012-002056). male male bleeding which might be associated with older
PPIs(inverse)andNSAIDscorrelatedwithre-bleeding
inwomen(R =−0.8754,P =0.002and ageandhighersalesofNSAIDs.
Received4September2012 male male ▪ Theresultshighlighttheneedforcarefulconsid-
R =0.7161,P =0.03,respectively),butnotin
Revised31October2012 female female eration of drug prescriptions to patients with
Accepted5November2012 men.AninversecorrelationbetweenPPIsalesand30-
daydeathafterbleedingwasfound(R =−0.9392, pepticulcerbleeding.
male
Thisfinalarticleisavailable Pmale=0.0002andRfemale=−0.8561,Pfemale=0.003),and Strengths and limitations
foruseunderthetermsof NSAIDsaleswerefoundtocorrelatewithincreased ▪ Drug sales/use and peptic bleeding outcomes
theCreativeCommons deathafterbleeding((Rmale=0.7278,Pmale=0.03, could not be linked with regard to individual
AttributionNon-Commercial Rfemale=0.7858,Pfemale=0.01). patients.
2.0Licence;see
Conclusions:ThesalesofNSAIDsandPPIscorrelate ▪ Re-bleedingcouldonlybeidentifiedonthebasis
http://bmjopen.bmj.com
withrecurrenceofpepticulcerbleedinginwomenand of re-admission, which means we might have
deathafterpepticulcerbleedinginbothgendersinthe lost information regarding re-bleeding that
populationlevel. occurredwithinthesamecaseofhospitalisation.
▪ Itisdifficulttofindasuitablecut-offdayfordef-
initionofre-bleeding.
INTRODUCTION Re-bleeding and death after peptic ulcer
Bleeding is one of the most frequent and bleeding occur in 7–16% and 3–14%,2 5 6
serious complications of peptic ulcer disease. respectively. These figures might increase as a
Despite the marked decrease in the occur- result of the increasing average age of many
rence of peptic ulcer disease during recent populations.1278Thehighriskofrecurrence
decades,parallelingthedecreasingprevalence and death highlights the need to identify the
Fornumberedaffiliationssee
of Helicobacter pylori infection, the incidence of best preventive measures available. The estab-
endofarticle
peptic ulcer bleeding has not changed appar- lished risk factors for peptic ulcer bleeding
ently. On the contrary, several surveys have include H pylori infection and medications
Correspondenceto
DrYunxiaLu; shownthattheincidenceofpepticulcerbleed- such as non-steroidal anti-inflammatory drugs
[email protected] ing has increased among older people.1–4 (NSAIDs), whereas proton pump inhibitors
LuY,SverdénE,LjungR,etal.BMJOpen2013;3:e002056.doi:10.1136/bmjopen-2012-002056 1
Drug sales/use and peptic ulcer bleeding
(PPIs) canpreventulcerbleeding.9Weaimedtoexamine diseases V.10 were used (K25.0, K25.4, K26.0, K26.4,
how the sales of PPIs and NSAIDs correlate with the inci- K27.0, K27.4, K28.0, K28.4, K92.0, K92.1 and K92.2).
dence, recurrence and death of peptic ulcer bleeding Since the treatment of ulcer perforation is different
fromapopulationperspective. from the treatment of ulcer bleeding, patients with per-
foration were excluded. Re-admission for peptic ulcer
bleeding within 1day of discharge was not regarded as a
METHODS new case of bleeding. Re-bleeding was defined as an
Study design episode of bleeding that occurred within 60days after a
This was a nationwide ecological study that addressed previous bleeding. Death was defined as any death
the correlation between relevant drug sales and peptic occurring within 30days of the date of admission for
ulcer bleeding in Sweden during the period 2000–2008. peptic ulcer bleeding. Death dates were obtained from
We used complete Swedish nationwide registers to the Death of Cause Register and the Swedish Population
collect data on sales of NSAIDs and PPIs, hospitalisation Register. The personal number, which is the unique
and death after peptic ulcer bleeding. The average daily identity for all the Swedish residents, was used to link
defined doses (DDDs) of NSAIDs and PPIs were com- dataamong different registers.
pared with the incidence, recurrence within 60days
after hospitalisation for bleeding, and 30-day death after
Statisticalanalyses
admission for peptic ulcer bleeding, in Sweden. The
Average DDD and time trends regarding the sales of
Regional Ethics Committee in Stockholm approved the
PPIs and NSAIDs were calculated on the basis of the
study.
average population for each year. DDD/TID was
described as DDDs/thousand inhabitants/day. A linear
Datacollection regression model was applied to test the statistical signifi-
Aggregated data on drug sales in Sweden during the cance of trends at the 5% level. Correlation analyses
study period were available from the Swedish Prescribed were performed between drug sales and the incidence,
Drug Register. This register records all prescribed and recurrence and death of peptic ulcer bleeding. All ana-
collected medications in the entire Swedish population lyses were gender-specific. Figures were plotted to show
of approximately 9million inhabitants.10 The Prescribed the correlations between drug sales and bleeding events.
Drug Register contains data on the age and sex of All statistical analyses were performed using SAS V.9.2
patientstogether withthe names ofprescribed drug sub- (SAS Institute, Cary,North Carolina, USA).
stances according to the anatomical therapeutic chem-
ical (ATC) classification. All NSAIDs (ATC codes: M01A)
and PPIs (ATC codes: A02BC and A02BD) were used for RESULTS
this study. All NSAIDs with ATC codes of M01A were Trends of PPI and NSAID sales
sold as prescription drugs except a few types of ibupro- The sales of PPIs increased during the study period,
fen inSweden. except for a temporary drop in 2003 (figure 1). The
Patients with peptic ulcer bleeding were identified sales of NSAIDs increased until 2004, after which there
from the Swedish Patient Register, which contains com- was a decrease to a level lower than in the year 2000
plete, nationwide data on all codes representing diagno- (figure 2). The decreased NSAID sales were particularly
ses and surgical procedures relating to inhospital care in evident in patients over 75years of age (data not
Sweden since 1987. Codes representing peptic ulcer shown). Women bought more PPIs and NSAIDs than
bleeding according to the international classification of men (figures 1 and 2), and this difference was more
Figure1 Salesofprotonpumpinhibitors(PPIs)indailydefineddoses(DDDs)/1000inhabitants/dayandhospitalisationratefor
pepticulcerbleeding(hospitalisationsforbleeding/100000inhabitants)inSwedenin2000–2008.(AandB)Representmenand
women,respectively.
2 LuY,SverdénE,LjungR,etal.BMJOpen2013;3:e002056.doi:10.1136/bmjopen-2012-002056
Drug sales/use and peptic ulcer bleeding
Figure2 Salesofnon-steroidalanti-inflammatorydrugs(NSAIDs)indailydefineddoses(DDD)/1000inhabitants/dayand
hospitalisationrateforpepticulcerbleeding(hospitalisationsforbleedingper100000inhabitants)inSwedenin2000–2008.(Aand
B)Representmenandwomen,respectively.
obvious with regard to NSAIDs. We also analysed the in men only (figure 2; R =0.6571, P =0.05,
male male
trend of sales of aspirin and H receptor antagonists P =0.2633,R =0.4177).
2 female female
(data not shown) which seems not very relevant with
trends of peptic ulcer bleeding, then we focused on cor- PPI and NSAID sales andpeptic ulcer re-bleeding
relations of PPIs/NSAIDs and peptic ulcer bleeding in The time trends of re-bleeding did not correlate with the
this study. sales of PPIs or NSAIDs in men (figures 3 and 4,
R =0.2227, P =0.5647; R =0.023, P =0.9522),
male male female female
Incidence, recurrence and death of peptic ulcer bleeding but the decreased occurrence of re-bleeding in women
correlated with the time trends of both PPI sales (R =
The hospitalisation rate for peptic ulcer bleeding was male
−0.8754, P =0.002) and NSAID sales (R =0.7161,
stable during the study period, although a higher rate male female
wasobservedinmenthaninwomen(figure1).Therate Pfemale=0.03).
of recurrence of bleeding was similar between the
PPI and NSAID sales and30-day death
genders, although the recurrence rate in women showed
a slightly decreasing trend (figure 3). Thirty-day death There was an inverse correlation between PPI sales and
after peptic ulcer bleeding decreased during the study death in both genders (Rmale=−0.9392, Pmale=0.0002,
period,especiallyinmen(figures5and6).Furthermore, Rfemale=−0.8561, Pfemale=0.003; figure 4), and the NSAID
women showed a higher death rate in different time salesshowedaclosecorrelationwithdeathofbleedingin
periods(figure6). both genders (Rmale=0.7278, Pmale=0.03, Rfemale=0.7858,
P =0.01;figure5).
female
PPIsand NSAIDs sales and peptic ulcer bleeding
Thetrendofpepticulcerbleedingdidnotcorrelatewith DISCUSSION
PPIs sales in either gender (figure 1; R =−0.2274, This study indicates that sales of NSAIDs and PPIs
male
P =0.5562,R =−0.2398,P =0.5342),butitcor- (inversely) correlates with 30-day death (both sexes) and
male female female
related marginally with the trend of the sales of NSAIDs recurrence (womenfemales) of peptic ulcer bleeding in
Figure3 Salesofprotonpumpinhibitors(PPIs)indailydefineddoses(DDDs)/1000inhabitants/dayandpepticulcer
re-bleedingratewithin60days(numberofre-bleeders/100inhospitalisationsforbleeding)inSwedenin2000–2008.(AandB)
Representmenandwomen,respectively.
LuY,SverdénE,LjungR,etal.BMJOpen2013;3:e002056.doi:10.1136/bmjopen-2012-002056 3
Drug sales/use and peptic ulcer bleeding
Figure4 Salesofprotonpumpinhibitors(PPIs)indailydefineddoses(DDDs)/1000inhabitants/dayand30-daydeathofpepticulcer
bleeding(numberofdeathswithin30days/100hospitalisationsforbleeding)inSweden2000–2008.(AandB)Representmenand
women,respectively.
the unselected population. Women appeared to have a reflux diseases. This might dilute the correlation
higher death of bleeding which might be associated with between PPIs and incidence of peptic ulcer bleeding.
ageand greater sales/use of NSAIDs. Nevertheless, we also have studied correlation between
The main strength of this study is the nationwide, PPIs/NSAIDs and re-bleeding, PPIs/NSAIDs and mortal-
complete data collection regarding drug sales, hospital- ity after bleeding diagnosis which may possibly further
isation for bleeding and death. Since there is virtually pinpoint the specific correlation between PPIs/NSAIDs
no private care for peptic ulcer bleeding in Sweden and andpepticulcerbleeding.Inaddition,afewtypesofibu-
since such bleeding usually requires hospitalisation, the profen as one of NSAIDs were sold as over-the-counter
incidence of bleeding and re-bleeding covered by this drugs which might lead to selection bias for NSAIDs in
study should represent population-based figures. There this study.Sincemost NSAIDs wereprescribed drugsand
are, however, several weaknesses that should be acknowl- actually few persons buy drugs without prescriptions in
edged. Drug sales/use and peptic bleeding outcomes Sweden due to the nationwide healthcare system, this
could not be linked with regard to individual patients. selectionbiasmaybenegligible.
This ecological design makes the interpretations more NSAIDs constitute an established risk factor for peptic
uncertain. Re-bleeding could only be identified on the ulcer bleeding, but the sales/use of these medications at
basis of re-admission, which means we might have lost the population level is less well-documented. A previous
information regarding re-bleeding that occurred within study reported that the sales of NSAIDs in Sweden
the same case of hospitalisation. On the otherhand, it is increased during the period 1978–2002.11 The present
difficult to find a suitable cut-off day for definition of study shows that sales have decreased since 2004 with
re-bleeding. The definition of death of bleeding was regard to both genders, and this decrease is more
based on death within 30days after discharge, since it is obvious in women and in the population aged over 75
difficult to assess if death actually results from bleeding. (data not shown). This decline might contribute to the
Furthermore, selection bias for PPIs users could exist decreasing trend of peptic ulcer death and re-bleeding
since PPIs were also prescribed for gastroesophageal in women. Similarly, the increased sales/use of PPIs
Figure5 Salesofnon-steroidalanti-inflammatorydrugs(NSAIDs)indailydefineddoses(DDDs)/1000inhabitants/dayand30-day
deathofpepticulcerbleeding(numberofdeathswithin30daysper100hospitalisationsforbleeding)inSwedenin2000–2008.(Aand
B)Representmenandwomen,respectively.
4 LuY,SverdénE,LjungR,etal.BMJOpen2013;3:e002056.doi:10.1136/bmjopen-2012-002056
Drug sales/use and peptic ulcer bleeding
Figure6 Salesofnon-steroidalanti-inflammatorydrugs(NSAIDs)indailydefineddoses(DDD)/1000inhabitants/dayandpeptic
ulcerre-bleedingratewithin60days(numberofre-bleeders/100hospitalisationsforbleeding)inSwedenin2000–2008.(AandB)
Representmenandwomen,respectively.
might contribute to these trends. The correlation might contribute to the significantly decreased occur-
between sales/use of PPIs and death indicates that PPIs rence of re-bleeding and death in this group in which
might prevent deaths from bleeding. This relation is PPI sales/use is increasing continuously.
more apparent in the male population, which shows a In conclusion, although the sales/use of NSAIDs and
very interesting, coinciding pattern. The different corre- PPIs in the general population does not seem to mirror
lations of drug and bleeding death between men and the incidence of peptic ulcer bleeding, such sales/use
women could be attributed to the different sales/use of correlate with re-bleeding and death of peptic ulcer
PPIs and NSAIDs between the genders. Previous studies bleeding. This correlation seems more obvious in old
have found that PPIs may help reduce peptic ulcer women which proposes an intriguing issue for future
bleeding, re-bleeding and the need for surgery, but not study. The potential reduced risk of death due to
death.12–14 It is difficult to differentiate between the decreased use of NSAIDs (especially in old women) and
respective roles of PPIs and NSAIDs. Nevertheless, the increased use of PPIs (especially in men) warrants
gastric mucosal protective effects of PPIs could be bene- further investigation.
ficial if NSAIDs are used. Surprisingly, PPIs are used in
only 10–22% of patients admitted with a peptic ulcer Authoraffiliations
disease and using NSAIDs.15–17 Our recent study, using 1UpperGastrointestinalResearch,DepartmentofMolecularMedicineand
Surgery,KarolinskaInstitutet,Stockholm,Sweden
individual prescription data from Sweden, revealed that
2SectionofUpperGastrointestinalSurgery,DepartmentofSurgery,South
lessthan 40% long-term NSAIDs users (those prescribed GeneralHospital,Stockholm,Sweden
NSAIDs for more than 180days) were prescribed con- 3CentreforEpidemiology,TheNationalBoardofHealthandWelfare,Stockholm,
comitantly gastroprotective medication.18 Furthermore, Sweden
4King’sCollegeLondon,London,UK
this study found that men took less gastroprotective
drugs compared with women, despite the fact that con- Contributors YL,RLandJLwereresponsibleforthedesignofthestudy.
RLwasresponsibleforacquisitionofthedata.YLwasresponsiblefordata
comitant antithrombotic treatment was more common
analysisincludingarrangementoftablesandgraphs.ES,CS,RLandJL
in men. On the other side, a German study found that
providedinputintodataanalysisandalltheauthorscontributedto
overprescription of drugs might increase risk of adverse interpretationoftheresults.YLinitiatedthedraftofthemanuscriptandthen
effect because of drug–drug interactions.19 It may indi- circulateditrepeatedlyamongalltheauthorsforcriticalrevision.
cate that the use of NSAIDs in high-risk population, for
Funding FundedbytheSwedishResearchCouncilandtheAstridandDavid
example, patients with a history of complicated peptic HagelénFoundation.
ulcer and elderly population with several concomitant
Competinginterests None.
prescriptions, should be cautious, and a personalised
Ethicsapproval TheRegionalEthicsCommitteeinStockholm.
regimen should beconsidered thoroughly.
The incidence of peptic ulcer bleeding is higher in Provenanceandpeerreview Notcommissioned;externallypeerreviewed.
men than in women, but few studies have reported Datasharingstatement ThedatabasewasorderedfromtheSwedish
gender distribution with regard to re-bleeding and NationalBoardofHealthandWelfare.Thedataaresensitiveandwecannot
death. One British case series study suggested that sharethedataonthewebsite.ThedatawillbestoredinKarolinskaInstitutet
formorethan5yearsforfurthercheck-up.Ifanyoneisinterestedinthedata,
women are at a higher risk of perforation or death than
itmightbepossibletoapplyfromhttp://www.socialstyrelsen.se/.
men.20 The older mean age in women was, however, an
important factor in that study, and this is consistent with
our data (data not shown). The greater sales/use of REFERENCES
NSAIDs in women cannot be ignored. The rapid 1. BlatchfordO,DavidsonLA,MurrayWR,etal.Acuteupper
gastrointestinalhaemorrhageinwestofScotland:case
decrease of NSAID sales/use in women, since 2004, ascertainmentstudy.BMJ1997;315:510–14.
LuY,SverdénE,LjungR,etal.BMJOpen2013;3:e002056.doi:10.1136/bmjopen-2012-002056 5
Drug sales/use and peptic ulcer bleeding
2. vanLeerdamME,VreeburgEM,RauwsEA,etal.AcuteupperGI 11. HermanssonM,EkedahlA,RanstamJ,etal.Decreasingincidence
bleeding:didanythingchange?Timetrendanalysisofincidenceand ofpepticulcercomplicationsaftertheintroductionoftheproton
outcomeofacuteupperGIbleedingbetween1993/1994and2000. pumpinhibitors,astudyoftheSwedishpopulationfrom1974–2002.
AmJGastroenterol2003;98:1494–9. BMCGastroenterol2006;25:CD002094.
3. LongstrethGF.Epidemiologyofhospitalizationforacuteupper 12. LeontiadisGI,SharmaVK,HowdenCW.WITHDRAWN:proton
gastrointestinalhemorrhage:apopulation-basedstudy.AmJ pumpinhibitortreatmentforacutepepticulcerbleeding.Cochrane
Gastroenterol1995;90:206–10. DatabaseSystRev2006;25:CD002094.
4. YavorskiRT,WongRK,MaydonovitchC,etal.Analysisof3294 13. RayWA,MurrayKT,GriffinMR,etal.Outcomeswithconcurrentuse
casesofuppergastrointestinalbleedinginmilitarymedicalfacilities. ofclopidogrelandproton-pumpinhibitors:acohortstudy.AnnIntern
AmJGastroenterol1995;90:568–73. Med2010;152:337–45.
5. VreeburgEM,SnelP,deBruijneJW,etal.Acuteupper 14. ZedPJ,LoewenPS,SlavikRS,etal.Meta-analysisofprotonpump
gastrointestinalbleedingintheAmsterdamarea:incidence, inhibitorsintreatmentofbleedingpepticulcers.AnnPharmacother
diagnosis,andclinicaloutcome.AmJGastroenterol 2001;35:1528–34.
1997;92:236–43. 15. DiFioreF,LecleireS,MerleV,etal.Changesincharacteristicsand
6. HearnshawSA,LoganRF,LoweD,etal.Acuteupper outcomeofacuteuppergastrointestinalhaemorrhage:acomparisonof
gastrointestinalbleedingintheUK:patientcharacteristics,diagnoses epidemiologyandpracticesbetween1996and2000inamulticentre
andoutcomesinthe2007UKaudit.Gut2011;60:1327–35. Frenchstudy.EurJGastroenterolHepatol2005;17:641–7.
7. PaspatisGA,MatrellaE,KapsoritakisA,etal.Anepidemiological 16. OhmannC,ImhofM,RuppertC,etal.Time-trendsinthe
studyofacuteuppergastrointestinalbleedinginCrete,Greece.Eur epidemiologyfpepticulcerbleeding.ScandJGastroenterol
JGastroenterolHepatol2000;12:1215–20. 2005;40:914–20.
8. BarkunA,SabbahS,EnnsR,etal.TheCanadianRegistryon 17. JohnellK,FastbomJ.Concomitantuseofgastroprotectivedrugs
NonvaricealUpperGastrointestinalBleedingandEndoscopy amongelderlyNSAID/COX-2selectiveinhibitorusers:anationwide
(RUGBE):Endoscopichemostasisandprotonpumpinhibitionare register-basedstudy.ClinDrugInvestig2008;28:687–95.
associatedwithimprovedoutcomesinareal-lifesetting.AmJ 18. LjungR,LuY,LagergrenJ.Highconcomitantuseofinteracting
Gastroenterol2004;99:1238–46. drugsandlowuseofgastroprotectivedrugsamongNSAIDusersin
9. LeontiadisGI,SharmaVK,HowdenCW.Systematicreviewand anunselectedelderlypopulation:anationwideregister-basedstudy.
meta-analysisofprotonpumpinhibitortherapyinpepticulcer DrugsAging2011;28:469–76.
bleeding.BMJ2005;330:568. 19. AhrensD,ChenotJF,BehrensG,etal.Appropriatenessof
10. WettermarkB,HammarN,ForedCM,etal.ThenewSwedish treatmentrecommendationsforPPIinhospitaldischargeletters.Eur
PrescribedDrugRegister—opportunitiesforpharmacoepidemiological JClinPharmacol2010;66:1265–71.
researchandexperiencefromthefirstsixmonths.Pharmacoepidemiol 20. SmartHL,LangmanMJ.Lateoutcomeofbleedinggastriculcers.
DrugSaf2007;16:726–35. Gut1986;27:926–8.
6 LuY,SverdénE,LjungR,etal.BMJOpen2013;3:e002056.doi:10.1136/bmjopen-2012-002056