Objective: To evaluate the effect of an educational interventionprogram on clinical outcomes and on compliance with medicaltherapy in patients with type 2 diabetes mellitus (DM), hypertension,or both.
Study Design: Six-month randomized unblinded study.
Know Your Health
Methods: Three hundred fifty-two patients were screened, and347 were randomized to the intervention group (education throughthe [KYH] program [n = 174]) or to the controlgroup (usual care [n = 173]). Evaluation of the effectiveness of theKYH program was based on the cohort of patients who were not atgoal at baseline (124 in the intervention group and 115 in the controlgroup). The primary research interests were to assess patientacceptance of the KYH materials and to compare the clinical outcomesof the intervention group with those of the control group.
Results: After 6 months, significantly more patients in the interventiongroup than in the control group were at goal (44.2% vs29.2%, = .046). Among patients with hypertension, reductions inthe mean diastolic blood pressure were significantly greater in theintervention group compared with the control group at month 6(-6.7 vs -3.6 mm Hg, = .04). The groups did not differ significantlyon other primary end points (percentage of patients with DMwho were at goal, change from baseline glycosylated hemoglobinlevel, and change in Morisky score).
Conclusions: Participation in the KYH educational program duringa 6-month period improved clinical outcomes in patients withtype 2 DM or hypertension. The KYH materials were well receivedand were considered informative and easily comprehensible bypatients who completed the program.
(Am J Manag Care. 2006;12:SP33-SP39)
Cardiovascular disease is the number one killer inthe United States.1 According to the AmericanHeart Association,1 cardiovascular disease is theunderlying cause of death in more than 910 000 adults,representing 37% of all deaths. Type 2 diabetes mellitus(DM) and hypertension are well-established risk factorsfor cardiovascular disease. An estimated 20 million persons(9.6% of adults) have DM, and 65 million persons(32.3%) have hypertension.1 Uncontrolled DM underminesemployee productivity and increases the use ofhealthcare resources (eg, hospitalization and physicianvisits) for the management of associated amputations,blindness, kidney disease, and nerve damage.2 The effectis compounded by restrictions on activities of daily livingand by a high incidence of depression.2 Compared withtheir peers without DM, adults with DM are more likelyto earn less, be unemployed, have more absences fromwork, and be limited in the type and amount of workthey can perform.2 By increasing the incidences of heartdisease, DM, stroke, kidney disease, depression, and lungdisease, uncontrolled hypertension also raises healthcareconsumption and costs, limits work, and contributesto coronary heart disease, the leading cause ofpremature permanent disability in the United States.3,4
Despite medications to control both disorders, nonadherenceto protracted therapy has long been recognizedas a major barrier to the optimum care of chronicdiseases like DM and hypertension.5-7 Adherence ratesof 50% to 75% of prescribed dosages may be typical.7-11Attempts to improve adherence have taken the form ofpatient education, written information, behavior modification,cash incentives, and directly observed therapy.12-15 Because education remains a cornerstone ofoptimal therapy, nonadherence may reflect inadequateor inappropriate education. This failure cannot beattributed solely to the patients and may serve toremind healthcare professionals that patients mustunderstand why, how, and when to follow therapeuticrecommendations. This study was performed to determinethe effect of an education program ([KYH]) on clinical outcomes and on compliancewith medical therapy among patients with DM, hypertension,or both in a large employer group.
Sixth Report of
the Joint National Committee on Prevention, Detection
Evaluation, and Treatment of High Blood Pressure
This 6-month randomized unblinded trial enrolled300 patients whose type 2 DM or hypertension wasuncontrolled according to criteria of the ,(blood pressure, <140/90 mm Hg)16 and the AmericanDiabetes Association (glycosylated hemoglobin [A1C]level, <7.0%).17 Screening sessions took place atLockheed Martin's Georgia and Mississippi sites to identifythe targeted patient population. All patients had adiagnosis, were in treatment with primary care physicians,and were being followed up through LockheedMartin's tertiary prevention program, which assists withtreatment compliance. Before randomization, thosepatients who were not at goal had baseline evaluationsperformed by the study team regarding blood pressure,A1C level, or both to determine if they still met theinclusion criteria. A randomization sequence was computergenerated for each patient; once study eligibilitywas determined, patients received the next availablerandomization number. Approximately 150 patientseach were to be randomized to the intervention groupand to the control group. Eligible patients received aneducational intervention (intervention group) or usualcare (control group) during 6 months of follow-up. Nomedications were provided, and all participants in bothgroups were encouraged to see their physicians onenrollment. At the 3-month and 6-month follow-upvisits, blood pressure readings and A1C levels wererecorded by the study team. Before patient enrollment,the protocol was reviewed and approved by the WesternInstitutional Review Board, Olympia, Washington.Written informed consent was obtained before thescreening procedures were completed.
Eligible patients were at least 18 years old with anestablished diagnosis of type 2 DM, hypertension, orboth. Diabetes mellitus was defined as a random plasmaglucose level of 200 mg/dL or higher (<11.1 mmol/L)with clinical symptoms, a fasting plasma glucose level of126 mg/dL or higher (<7.0 mmol/L), or an A1C levelgreater than 7.0%. Hypertension was defined as a systolicblood pressure of 140 mm Hg or higher or a diastolicblood pressure (DBP) of 90 mm Hg or higher.Patients were required to speak English, not be pregnantor lactating, and be cognitively aware without a diagnosisof dementia or organic brain syndrome. Failure tomeet these criteria was reason for exclusion, as wererefusal to participate and absence of baseline values.The effectiveness of the intervention was based on evaluationsat baseline, at month 3, and at month 6.
The present study was modeled after the ongoingFlorida Health Literacy Study, which is using the KYHprogram in a community health center. The KYH programintegrates established culturally sensitive healtheducation practices for self-management of type 2 DMand hypertension with communication strategies andtechniques designed specifically for populations withlow functional health literacy. The KYH program in thepresent study provided 1 hour of education for patientswith hypertension and 3 hours of education for patientswith DM and was conducted by trained facilitators onLockheed Martin property. Participants were encouraged,but not required, to enroll in the on-site fitnesscenter, and any and all physical activity was promoted.Individual programs were not monitored. The programalso provided patients with diet and exercise regimensand with tools to track visits to their physicians.
The statistical analysis plan was finalized beforedatabase quality control and evaluability assessmentswere made and statistical analyses were conducted.The primary analysis cohort included all patients whowere not at goal for hypertension, DM, or both atbaseline and who had at least 1 postbaseline visit.Analyses were conducted at the month 3, month 6,and end point visits (end point was taken as the lastpostbaseline assessment for each patient). Secondaryanalyses were conducted on the subsets who had DMor hypertension. Participants with dual diagnoses wereincluded in both subsets. Categorical measures, suchas the number of patients at goal, were assessed usingthe general association variant of the Cochran-Mantel-Haenszel c2 test. At each time point, analyses ofcovariance were applied to continuous measures, suchas the changes from baseline in blood pressure. To provideadditional statistical power, changes from baselinein A1C level and in systolic blood pressure and DBP atthe 3-month and 6-month evaluations were analyzedusing linear contrasts with repeated-measures random-effects models (PROC MIXED, SAS softwareversion 8; SAS Institute, Cary, NC). Descriptive statisticswere calculated for all primary and secondarymeasures at baseline, at 3 months, and at 6 months.All statistical tests were 2-sided, and significance wasset at = .05. Enrollment of approximately 150 patientswith uncontrolled disease per group was designed toprovide 80% power, with a type I error rate of .05, todetect a 16% between-group difference in the percentageof patients at goal. The calculation was based on anassumed 30% response rate in the control group (46%in the intervention group).
At baseline, 352 patients were screened for enrollment,and 347 were eligible for randomization. Of thosepatients not at goal, 124 were assigned tothe intervention group and 115 to thecontrol group (the efficacy-evaluablecohorts). Three patients were not evaluatedat any visits and were excluded fromthe analysis. The study groups werewell matched for demographics, workenvironment, general health, and responsesto baseline questions (Table 1).Most patients in both groups werewhite or African American men. Theirage ranged from 22 to 80 years, andmost participants had at least somecollege or technical school education.Almost all patients worked full-time onthe first shift, with about 50% in manufacturing.Annual salaries averagedbetween $50 000 and $75 000. Approximately90% of both groups describedthemselves as nonsmokers.Participants generally reported thatthey were recreational exercisers whorated their health as good and had nohealth-related limitations in carryingout moderate activities or in climbingstairs. The Short Test of FunctionalHealth Literacy in Adults scores indicateda more than adequate literacylevel in 98.6% of patients.
At baseline, 24 patients (6.9%) had DMonly, 229 (66.0%) had hypertension only,and 91 (26.2%) had a dual diagnosis(Figure 1). Overall, 23.6% of the interventiongroup and 17.6% of the control groupleft the study prematurely, with 76.4% and82.4%, respectively, completing the study.Between 92.9% and 100.0% of patientswith DM tested their blood glucose levelregularly (5-6 days/wk) (Table 2).Between 56.4% and 60.9% of patientswith hypertension tested their blood pressureregularly (5-6 times/mo). Morepatients with DM (85.7%-95.2%) took medication thanpatients with hypertension (67.4%-69.1%). Substantiallymore patients with DM (57.1%-81.0%) reported havingreceived education about their condition comparedwith patients with hypertension (13.0%-13.8%).
After randomization to the intervention group, thepercentage of patients in the efficacy-evaluable cohortwho reached goal was 30.0% at month 3, with afurther increase to 44.2% at month 6 (Table 3 andFigure 2). The proportion of the control group thatreached goal after randomization was greater atmonth 3 (33.3%), dropping to 29.2% at month 6. Thedifference in goal achievement between the groups wasstatistically significant in favor of the intervention atmonth 6 (= .046).
Changes From Baseline in GlycosylatedHemoglobin Level (Patients With DM)
There were no statistically significant between-groupdifferences for changes in A1C level from baseline tomonth 3 or month 6 (Table 3). The repeated-measuresanalysis also showed no significant differences in thisvariable, primarily because of the small number of patientswith DM enrolled.
Changes From Baseline in Blood Pressure(Patients With Hypertension)
The mean systolic blood pressure and DBP decreasedfrom baseline to the postbaselineevaluations in the interventiongroup and in the control group(Table 3). These decreases wereconsistently larger in the interventiongroup at month 3 and atmonth 6. The difference in DBPreduction at month 6 was significantlygreater in the interventiongroup (= .04) based on a repeated-measures analysis.
In both groups, mean Moriskyscores indicated a shift towardhigher-level compliance behaviorat month 3 and month 6, althoughno statistically significantdifferences between the groupswere seen (Table 3).
Readiness to ChangeQuestionnaire(DM Only)
The mean scores on theReadiness to Change Questionnaireincreased relative to baselinein both groups at month 3and month 6 (Table 3). The differencebetween the groups wasnot statistically significant.
Patient SatisfactionQuestionnaire(Intervention Group Only)
In the intervention group,83.3% of patients with DM and86.6% of patients with hypertensionwere very satisfied with theKYH intervention program. Mostpatients, including those withthe highest levels of education,rated the program as 4 or 5 on a5-point scale. Most patients alsoexpressed appreciation thattheir company provided the program and indicated thatthey would attend similar learning programs coveringother diseases.
Changes in Additional Measures
The groups did not differ significantly in thechange from baseline body mass index. By month 3according to a questionnaire on patient characteristics, 73.3% of the interventiongroup had seen theirphysicians since their lastvisit compared with 63.3%of the control group. Thisbehavior allowed moreintervention patients thancontrol patients to changetheir regular medication(47.8% vs 31.1%, = .02).
Systematic interventionwith a program that integratesestablished culturallysensitive health educationpractices for self-managementcan help improve clinicaloutcomes in patients withtype 2 DM and hypertension. In this study, statisticallysignificant differences in the primary end points wereobserved after implementation of the KYH program.Significantly more patients in the intervention groupwere at goal after 6 months compared with the patientsin the control group (44.2% vs 29.2%, = .046).Although the mean blood pressure decreased in bothgroups, the intervention group demonstrated a significantlygreater mean reduction in DBP compared with thecontrol group at month 6 (-6.7 vs -3.6 mm Hg, = .04).No statistically significant between-group differences wereobserved on other primary end points (percentage ofpatients with DM who were at goal, change from baselineA1C level, and change in Morisky score).
The KYH program was well received by patientsassigned to the intervention group: 76.4% attended theeducational session and more than 80% reported beingsatisfied with the program, indicating acceptance of theprogram and its benefits. This comprehensive programwas designed to help patients with DM and hypertensiontake better care of themselves by educating them abouttheir medical condition, the importance of a healthfuldiet and exercise, the need to check blood glucose leveland blood pressure regularly and frequently, and thebenefits of discussing these issues with their physicians.At the conclusion of the program, almost all participantsfelt healthier and in better control of their health. Evenhighly educated participants accepted the comparativelylow educational level of the program materials.
The substantial blood pressure reductions in theintervention group likely reflect the effect of educationand the reinforcement of adherence.The educational focus on understandingall relevant risk factors used a structuredapproach that emphasized theimportance of lifestyle changes, followupwith physicians, regular blood pressuremonitoring, and adherence tostudy and nonstudy interventions,including medication use. The positivemessages of the KYH program empowerpatients to reduce their risks by makingfeasible changes. The finding of substantial,albeit smaller, reductions in bloodpressure in the control group suggeststhat enrollment in the study andincreased interaction with healthcareprofessionals may have improved theadherence to treatment. Supporting thishypothesis is the fact that adherence inthe control group improved as much asthat in the intervention group, an effectobserved in earlier clinical trials.18-20
The absence of effect on A1C level in patients with DMmay have at least 2 explanations. First, the small numberof these patients may have limited the statistical power todetect differences during follow-up. Second, the patientswith DM in both study groups were at substantially higherlevels of awareness and functioning about their disorderon entry than the patients with hypertension. For example,many more patients with DM had already been educatedabout their disease, were monitoring glucose levelsregularly, and were taking appropriate medication.Patients with DM in the intervention group and in thecontrol group may have already been taking better care ofthemselves and had less room for improvement.
In conclusion, the primary efficacy analysis of thisstudy demonstrated that participation in the comprehensiveKYH program significantly improved attainmentof treatment goals, with a significant reduction in DBP inpatients with hypertension. The KYH materials werewell received and were considered informative and easilycomprehensible by the study participants.
From Wellness and Health Promotion, Lockheed Martin, Marietta, Ga (PDT); andGlobal Statistics, Pfizer, Inc, New York, NY (RM).
This study was funded by Pfizer, Inc. Funding included compensation for an independentcontractor to collect and analyze the data, to ensure compliance with institutionalreview board requirements, and to prepare a manuscript. Pfizer, Inc had the right to approveor disapprove publication of the final manuscript.
Address correspondence to: Robert Miceli, PhD, Global Statistics, Pfizer, Inc, 235 E42nd St, New York, NY 10017. E-mail: firstname.lastname@example.org.
Heart Disease and Stroke Statistics: 2006 Update.
1. American Heart Association. Dallas, Tex: American Heart Association; 2006.
2. National Academy on an Aging Society. Diabetes: a drain on U.S. resources.April 2000. Available at: http://www.agingsociety.org/agingsociety/pdf/diabetes.pdf.Accessed November 2, 2006.
3. National Academy on an Aging Society. Hypertension: a common condition forolder Americans. October 2000. Available at: http://www.agingsociety.org/agingsociety/pdf/hypertension.pdf. Accessed November 2, 2006.
4. National Academy on an Aging Society. Heart disease: a disabling yet preventablecondition. January 2000. Available at: http://www.agingsociety.org/agingsociety/pdf/heart.pdf. Accessed November 2, 2006.
BMC Fam Pract.
5. Littenberg B, MacLean CD, Hurowitz L. The use of adherence aids by adultswith diabetes: a cross-sectional survey. 2006;7:e1.
Ann Intern Med.
6. Rodondi N, Peng T, Karter AJ, et al. Therapy modifications in response to poorlycontrolled hypertension, dyslipidemia, and diabetes mellitus. 2006;144:475-484.
7. Perreault S, Lamarre D, Blais L, et al. Persistence with treatment in newly treatedmiddle-aged patients with essential hypertension. 2005;39:1401-1408.
in Health Care.
8. Haynes RB. Introduction. In: Haynes RB, Taylor DW, Sackett DL, eds. Baltimore, Md: Johns Hopkins University Press; 1979:1-7.
9. Schectman JM, Bovbjerg VE, Voss JD. Predictors of medication-refill adherencein an indigent rural population. 2002;40:1294-1300.
10. Cramer JA. A systematic review of adherence with medications for diabetes.2004;27:1218-1224.
11. Cramer JA. Effect of partial compliance on cardiovascular medication effectiveness.2002;88:203-206.
12. Roter DL, Hall JA. Strategies for enhancing patient adherence to medical recommendations.1994;271:80.
Adherence to Treatment in Medical Conditions.
13. Warren L, Hixenbaugh P. Adherence and diabetes. In: Meyers LB, Midence K,eds. Amsterdam, the Netherlands:Harwood Academic Publishers; 1998:423-453.
Bull World Health Organ.
14. Chee CB, James L. The Singapore Tuberculosis Elimination Programme: the firstfive years. 2003;81:217-221.
Int J Tuberc Lung Dis.
15. White MC, Tulsky JP, Reilly P, McIntosh HW, Hoynes TM, Goldenson J. A clinicaltrial of a financial incentive to go to the tuberculosis clinic for isoniazid afterrelease from jail. 1998;2:506-512.
Arch Intern Med.
16. Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation,and Treatment of High Blood Pressure. 1997;157:2413-2442.
17. American Diabetes Association. Standards of medical care in diabetesâ€”2006.2006;29(suppl):S4-S42.
Ann Intern Med.
18. Roumie CL, Elasy TA, Greevy R, et al. Improving blood pressure controlthrough provider education, provider alerts, and patient education: a cluster randomizedtrial. 2006;145:165-175.
J Hum Hypertens.
19. Roca-Cusachs A, Sort D, Altimira J, et al. The impact of a patient educationprogramme in the control of hypertension. 1991;5:437-441.
West Indian Med J.
20. Duff EMW, Simpson SH, Whittle S, Bailey EY, Lopez SA, Wilks R. Impact onblood pressure control of a six-month intervention project. 2000;49:307-311.