Research ArticleOutcome Analysis of Hemoglobin A1c, Weight, and
BloodPressure in a VA Diabetes Education ProgramSusan L. North, RD,
MPH; Glen A. Palmer, PhD, ABNABSTRACTObjective:
Todeterminetheeffectofaspecicdiabeteseducationclass(Basics)onhemoglobinA1cvalues,
weight, and systolic blood pressure.Design:
Inthisretrospectivestudy,
theresearcherscompared2groupsofmaleveteranswitharecentdiagnosisoftype2diabetes.
Onegroupreceiveddiabetesgroupeducation(n 175)overa4-monthperiod,
and the other received standard diabetes management follow-up (n
184).Setting: Outpatient clinic setting in the
Midwest.Interventions: Basics class compared with standard level of
care.MainOutcomeMeasures:
Pre-andpost-laboratoryvaluesforhemoglobinA1c,weight,andsystolicblood
pressure.Analysis: Multivariate analysis of covariance and
follow-up univariate statistics for signicant differences.Results:
Findings revealed signicant differences in hemoglobin A1c (P
4months' follow-up. Theyalsonoted that HbA1c decreased morewith
additional contact time betweenparticipantandeducator. Adecreaseof
1% was noted for every additional23.6hours of contact. One
percentdrop in HbA1c is associated withimproved outcomes.7Ina
meta-analysis of 153studiespublished between 1977 and
1994evaluatingtheeffectiveness of inter-ventions designed to
improve patientcompliance with medical
regimen,Roteretal8concludedthatnosingleintervention strategy
appeared consis-tentlystrongerthananyother.How-ever, the authors
noted that the morecomprehensivetheprogramwas,
themoreeffectivetheoutcomewas, andthe most benecial
interventionsincluded3comprehensiveelements:educational,
behavioral, and affective.Offering
groupeducationtopatientsmayprovetobemorecost-effective.9Duncanet
al10reviewedcommercialand Medicare claims
payer-deriveddatasetsandconcludedthatpatientswho participated in
diabetes educationhad lower average costs thanthosewho didnot
participate indiabeteseducation. In addition, diabetes
educa-tionparticipants weremorelikelytofollow recommendations for
care thansimilar patients who did not participatein diabetes
education. The authorsconcludedthatqualityofcarecanbeimproved and
costs reduced, specif-ically among men and people in disad-vantaged
areas.TheBasicsprogramisusedattheVAhealthcare systemfromwhichthis
retrospective chart review wasconducted. This curriculum
originatesfromoneof the rst outpatient dia-beteseducationcenters.
Clinical andeducational guidance was
providedbyaregistereddietitian, aregisterednurse, andaphysician.
Thephiloso-phy at the heart of the program is
therecognitionthatthepatientisatthecenterofthisteamandcanlearntoself-manage
diabetes. Four educationalmodels are applied to the
curriculum:adult learning theory,11the trans-theoretical model
(stages of change),12the healthbelief model,13and the
pub-lichealthnursingmodel.14Thepro-grammeets the National
StandardsforDiabetesSelfManagementEduca-tion Programs.15A detailed
descriptionand philosophy of the program can befound in Diabetes
Basics: education,innovation, revolution.16The
targetaudienceforthecurriculumisnewlydiagnosed persons with type 2
diabeteswhohave receivedlittle or noself-management education. The
objec-tivesoftheprogramarethreefold:toimprove glycemic control, to
enhanceknowledge of diabetes management,and, for patients, to
implement at least1 positive behavior change.In the mid 1990s, a
team from theInternational Diabetes Center in Min-neapolis, MN,
conducteda randomizedcontrolledtrial to compare diabetesgroup
education using the Basics cur-riculum(n 87)withindividual
in-struction(n 83). Subjects receivededucation in 4 sequential
sessionsover 6 months. Outcomes reviewedincluded change in
knowledge,self-management behaviors, weight,BMI, HbA1c,
health-related qualityoflife, patient attitudes, and
medicationregimen. Results showedsimilar im-provement inknowledge,
BMI, andhealth-related quality of life. Hemoglo-bin A1c decreased
in both groupsalthough the decrease was marginallygreaterin
groupeducationthanindi-vidual instruction. This study demon-strated
that group education andindividual education are equally effec-tive
at providing the outcome ofimproved glycemic control.17In this
particular VA outpatientclinic setting, veterans are
typicallyreferred to classes by a diabetes
educa-tionconsultplacedbytheirprimarycareprovider.
TheBasicsprogramispresented by a registered
dietitianandaregisterednursein3sessions.Sessions 1 and 2 are
planned forapproximately 2 weeks apart to
allowpatientstodeneandfocustheiref-forts for behavior change.
Session3is held3months after
session2toreneandpracticeeffortsfurtherto-wardlong-termbehavior
change inthose areas the patient has identied.Figure 1 shows the
content of each in-dividual session. The three sessionlengths are
2.5, 2, and 1.5 hours.Before each session, vital signs aretaken and
recorded in the medical re-cord. The dietitianweighs eachpa-tient
ona calibratedscale. Patientsareweighedfullyclothed.Thenursetakes a
bloodpressure reading beforeclass. An elevated blood
pressure(>140/90) is rechecked before theend of the
session.Althoughtheprimarygoal oftheBasics programis glycemic
controlfor the patient, management ofweight
andbloodpressuremayalsoprovide signicant health benets.Atracking
and goal setting toolcalledMySuccess Plan is
usedintheBasicscurriculumforthebenetof instructors and patients.
With thistool, instructors guide patients tofocus on1 area of
change from7deemed critical by the American Asso-ciation of
Diabetes Educators. These 7areas include nutrition, physical
activ-ity, medication regimens, problemsolving, stress management,
bloodglucose testing, andother forms ofrisk reduction such as
smoking cessa-tion. Patientsmaybenetbysettingmore measurable
andspecic goalsfor behavior change, thus avoidingfrustration and
confusion. Successplans helppatients prioritizeeffortsand dene
expectations.Family and peer support are recog-nized as important
to a patient'sself-management efforts. Therefore,veterans are
encouraged to bring afamilymember suchas aspouse toparticipate in
all sessions. Veteransalsomay learnfromeachother bysharing personal
struggles and suc-cesses. A regularly occurring follow-upgroup is
offered to veterans who havecompleted the 3 sessions. Follow-up
isencouragedquarterly, biannually, orannually per the veteran's
choice.Overall, basedontheliteraturere-view, diabetes
groupeducationmayresult in modest improvement inHbA1c, weight,
blood pressure, andother indicators of health. The primaryobjective
of this retrospective chartJournal of Nutrition Education and
BehaviorVolume 47, Number 1, 2015 North and Palmer 29reviewwas
todetermine the associationof diabetes education classes using
theBasicscurriculumwithbloodglucosemanagement, weight, and blood
pres-sure outcomes. It was hypothesizedthat individuals who
received diabeteseducation using the Basics
curriculumwouldhaveimprovedoutcomesoverindividuals who did not
participate inthese classes.METHODSThis study was a retrospective,
obser-vational chart review of medicalrecords froma large VA
medical centerintheMidwest that uses theBasicsprogram. The study
was approvedby the VA-afliated
institutionalreviewboardandresearchanddev-elopmentcommittee.
Becauseoftheretrospective nature of the study, theresearchers
obtained a waiver ofinformedconsent andHealthInsur-ance Portability
and AccountabilityAct authorizationfromthe afliateinstitutional
review board.Patient SampleThis was a retrospective analysis
ofmedical records of veterans whoreceived care from a large VA
medicalcenter inthe Midwest. Medical re-cords wereincludedinthis
studyifthe medical record showed a diagnosisof type 2 diabetes
within the previous2years.Ofthe1,626patientrecordsscreened, 1,263
were initiallyexcluded in the study. Four
additionalpatientrecordswereexcludedowingtoextremeoutliers
onpretreatmentscores. Themajorityof screenedpa-tient records that
were excludedfromthis study did not meet
thecriteriaforarecentdiagnosisoftype2 diabetes or did not have
laboratoryvalues that fell within establishedtime parameters.
Figure 2 shows aowchart of the selection process.A total of 359
medical records(Basics class 175 and controlgroup 184)
wereacceptedfor thisstudy. The time period for this chart re-view
is between the beginning ofJanuary, 2008andtheendof June,2011.
Treatment group candidateswere identied froma spreadsheetkept
internally by the medical center'snutrition clinic, which tracks
pre- andpost-class values for HbA1c, weight,andsystolic
bloodpressure andhasbeenusedforcasemanagementpur-poses within the
facility. Each patientinthetreatmentgroupcompletedall3 sessions of
the Basics diabetes
educa-tionprogram.Thecontrolgroupwasselectedfromalist of veterans
whoare patients of the medical center,whohavea diagnosis of
type2diabetes,and had not beenseen ina VAdiabeteseducation
clinic.The mean age of individuals in theBasics class was 65.2
years (SD, 8.87);the meanage of the control groupwas 66.8 (SD,
9.67). Most of thesubjects were Caucasian(78.9%forthe treatment
group vs 73.4% for thecontrol group). There were no signi-cant
differences between groups basedon age (P .11) or ethnicity (P
.33).Chart ReviewFor the treatment group, the re-searchers
collected data includingweight and systolic blood
pressurereadingsthatareroutinelytakenandrecordedbeforesessions1and3andwere
present in the medical record. If2 systolic blood pressure
readingswere present in the medical record forthe same class day,
the lower of the 2was taken. For thetreatment group,HbA1c values
were taken fromtheclosest availablereadingbeforetreat-ment
(nogreater than90days) andthe earliest reading posttreatment
(nogreater than 90 days). Total length oftime between pre- and
posttreat-mentwasnomorethan10months.This period was determined to
beappropriate because the accuracy ofan HbA1c value is
approximately6090 days. A medical record wasincludedinthestudyif
theveteranhadbeengivenadiagnosisoftype2diabeteswithintheprevious2years.The
researchers collected data forthe control groupincluding
HbA1c,weight, and systolic blood pressurefound in the medical
record at baselineandfollow-upforthesamelengthoftime as the
treatment group. Data pre-sent in the medical records wereassumed
to have been collectedfollowing standard VA procedure.Figure1.
Educational topicspresentedintheBasicsprogrambysession.30 North and
Palmer Journal of Nutrition Education and BehaviorVolume 47, Number
1, 2015Statistical AnalysisAminimumsamplesizeof 211wasdetermined as
necessary to provideforapowerof0.80andeffectsizeof0.25. Alpha for
this calculationwassetat P