N/A = Not Available or Not Applicable NOS = Not Otherwise Specified * Unless otherwise specified, Mean (Standard Deviation) reported. 156 Evidence Table 1: Diabetes First Author Year (ID) Condition (Type) Study Design Quality Population Characteristics Arm Intervention Sample Size Intervention Characteristics Meta-Analysis Data* or Outcomes Follow-up Time(s) 1 Dietary monitoring (Office visit) Education (Office visit) Education (One-on-one) Education (Reading material) Exercise diary (Office visit) Feedback (Office visit) Practice methods (Protocols) n Entered: n/a n Analyzed: 27 Tailored: Yes Group Setting: No Feedback: Yes Psychological: No Primary MD: n/a Allen BT, 1990 (#2201) Diabetes (Type II) RCT Jadad Score: 3 Diagnostic criteria: FBS Comorbidities: Obesity and cholesterol 2 Dietary monitoring (Office visit) Education (Office visit) Education (One-on-one) Education (Reading material) Exercise diary (Office visit) Feedback (Office visit) Practice methods (Protocols) n Entered: n/a n Analyzed: 27 Tailored: Yes Group Setting: No Feedback: Yes Psychological: No Primary MD: n/a Excluded from meta-analysis as no usual care or comparable control group. Patients who self monitored diabetes using urine testing (arm 1) had similar statistically significant reductions in fasting blood glucose, glycosylated hemoglobin, and weight as did patients utilizing serum glucose testing (arm 2). No appreciable differences between groups were noted. Follow-up times: 1 MO, 2 MO, 3 MO, 4 MO, 5 MO, 6 MO 1 Usual Care (n/a) n Entered: n/a n Analyzed: 23 Tailored: n/a Group Setting: n/a Feedback: n/a Psychological: n/a Primary MD: n/a Anderson R M, 1995 (#747) Diabetes (n/a) CCT Jadad Score: 0 Diagnostic criteria: n/a Comorbidities: n/a 2 Education (Group meeting) Education (Video/audio tapes) Feedback (Group meeting) n Entered: n/a n Analyzed: 22 Tailored: Yes Group Setting: Yes Feedback: Yes Psychological: No Primary MD: n/a Excluded from meta-analysis as not randomized. Patients receiving a patient empowerment education program (arm 2) had reductions in glycosylated hemoglobin that were greater than controls and were statistically significant (p=0.05). Intervention subjects also improved in all self-efficacy sub-scales, which were sustained at 12-week follow-up. Follow-up times: 6 WK, 12 WK
83
Embed
Evidence Table 1: Diabetes - Home - Centers for Medicare ... · Evidence Table 1: Diabetes First Author Year (ID) Condition (Type) Study Design Quality Population Characteristics
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
156
Evidence Table 1: Diabetes
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Patients who self monitored diabetes usingurine testing (arm 1) had similar statisticallysignificant reductions in fasting blood glucose,glycosylated hemoglobin, and weight as didpatients utilizing serum glucose testing (arm 2).No appreciable differences between groupswere noted.
Patients receiving a patient empowermenteducation program (arm 2) had reductions inglycosylated hemoglobin that were greater thancontrols and were statistically significant(p=0.05). Intervention subjects also improved inall self-efficacy sub-scales, which weresustained at 12-week follow-up.
Follow-up times: 6 WK, 12 WK
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
157
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Patients randomized to intervention (arm 2) hada greater number of MD evaluations but nodifference in diabetes related hospitalizationscompared with controls (arm 1). BMI trendswere higher in intervention patients comparedwith controls, but there were no treatmentdifferences in glycosylated hemoglobin, systolicor diastolic blood pressure. There were also nosignificant differences in diabetes knowledge,anxiety, depression, satisfaction with treatmentor self reported well-being.
Excluded from meta-analysis as no usual careor comparable control group.
Though knowledge and% ideal body weightsignificantly improved for Learning ActivityPackages (arm 1) at 5 months and HgbA1c andbehavior improved for diabetes class arm, onlyknowledge scores were significantly higher at 5months for the LAP arm.
Follow-up times: 2 MO, 5 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
158
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
InterventionSample Size
InterventionCharacteristics
Meta-Analysis Data* or OutcomesFollow-up Time(s)
1 Control (n/a)Advocacy training (One-on-one)Counseling/therapy (One-on-one)Education (Group meeting)Follow up (One-on-one)
Excluded from meta-analysis as no usual careor comparable control group.
Intervention subjects (arm 2) had greaterdecreases in HbA1c levels than those receivingusual care (arm 1) (1.7% versus 0.6% p<0.01).Fasting serum glucose was lower in interventionsubjects by a mean of 48 mg/dl versus 15 mg/dl(p=0.003). Self-rated health also improved inthe intervention group (p=0.02).
Follow-up times: 6 MO, 12 MO
1 Control (n/a)Education (One-on-one)Education (Reading material)
Tailored: YesGroup Setting: YesFeedback: NoPsychological: NoPrimary MD: No
Follow-up time not in 3 - 12 months.
Though subjects randomized to an educationintervention (arm 2) demonstrated increasedknowledge compared with usual care group(arm 1) (p=0.007) and had significant reductionsin HbA1c and fasting blood glucose, thesereductions in biochemical markers were notsignificantly greater than in the usual caregroup. There were also no changes incholesterol, blood pressure, or foot lesions andhealth service utilization was unaffected.
Follow-up times: 18 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
160
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: YesPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
Behavioral strategy interventions (arms 2, 3 and4) resulted in no differences in glycosylatedhemoglobin and weight loss betweenintervention and control groups.
Follow-up times: n/a
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
161
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: NoFeedback: YesPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
Control (arm 1) were more likely to have anincrease in intensity of diabetes treatment at 6-month follow-up (p=0.04) than interventionsubjects (arms 2, 3, and 4). Behavior program(arm 4) and group education (arm 2) patientshad greater improvement in knowledge scoresat 6-month follow-up, but differences were notsustained at 12 months. Greater reductions indiastolic blood pressure were seen for thoseattending behavioral interventions (p=0.02). Nodifference in change between groups occurredfor HbA1c, BMI, total cholesterol, or systolicblood pressure.
Follow-up times: 3 MO, 6 MO, 12 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
162
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Patients in both intervention group (arm 2) andcontrol group (arm 1) lost weight. Thoughcholesterol levels fell significantly in the low fatgroup (arm 2) (p<0.001), mean plasma glucoseand HbA1c remained unchanged.
Follow-up times: 6 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
163
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: NoFeedback: NoPsychological: NoPrimary MD: No
Follow-up time not in 3 - 12 months.
Intervention subjects (arm 2) had greaterknowledge (p<0.005) and lower glycosylatedlevels (10.4% versus 11.8%, p<0.05) than usualcare group (arm 1) did at 4-6 weeks after theprogram concluded. Change in body weight wasnot different between groups.
Tailored: YesGroup Setting: YesFeedback: NoPsychological: NoPrimary MD: No
Insufficient statistics for meta-analysis.
Individuals participating in 2 nutrition groups(arms 2 and 3) demonstrated decreased caloricintake compared with usual care group (arm 1).The addition of a social learning program (arm3) had a significant decrease in weight at 2-month follow-up. Intervention conditionsproduced a marginal improvement in fastingblood glucose (p<0.08).
Follow-up times: 2 MO, 2 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
165
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Patients who received a brief intervention (arm2) had no improvement in HbA1C at 3-monthfollow-up when compared with usual care group(arm 1). However serum cholesterol wassignificantly lower (p<0.001) in the interventiongroup as were 4 dietary behavioral measures.Though patient satisfaction was improved,quality of life was not.
Follow-up times: 3 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
166
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
InterventionSample Size
InterventionCharacteristics
Meta-Analysis Data* or OutcomesFollow-up Time(s)
1 Control (n/a)Clinical reviews w/patient (One-on-one)Education (Office visit)Education (Reading material)
Excluded from meta-analysis as no usual careor comparable control group.
Intervention subjects (arms 2 and 3) reportedgreater physical activity than controls (arm 1) at5-year follow-up (p<0.01). Intervention subjectsalso had better control of glucose and lowersystolic blood pressure (143 versus 154 mmHg,p<0.01) and required fewer antidiabetic drugs.Though no differences between groups werenoted for myocardial infarction incidence,cumulative incidence mortality rates suggesteda benefit from intervention.
Follow-up times: 2 YR, 5 YR
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
168
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Subjects who participated in a system of careshared between specialist and generalist (arm2) had significantly greater visit compliance thanthose with generalists alone (arm 3) (72%versus 35%, p<0.04). HbA1c improvedsignificantly in all 3 groups but no differencesbetween groups were noted. Nor were thereblood pressure differences between groups andweight decreased marginally in all 3 groupsthough this was statistically significant only inthe shared care group (arm 2) (p<0.04).
Follow-up times: 1 YR
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
169
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: YesPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
Those participating in the diet intervention (arm2) lost more weight than the other 3 groups(arms 1, 3, and 4) (p<0.05). HDL cholesterolwas also significantly higher in this group(p<0.01). No differences in glycosylatedhemoglobin between groups were noted.
Tailored: YesGroup Setting: YesFeedback: YesPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
At 18-months follow-up diabetic patientsreceiving behavioral interventions in a combineddiet and exercise program (arms 2, 3, and 4)achieved greater reductions in glycosylatedhemoglobin than those receiving only diet,exercise, or control interventions (arm 1)(p<0.05). Changes between other interventionswere not significant. Improvements in quality oflife measures were also greatest in thecombined group (arm 2) (p<0.05).
Follow-up times: 3 MO, 6 MO, 12 MO, 18 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
172
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Both diet guide (arm 1) and exchange liststreatment group (arm 2) demonstratedsignificantly higher levels of self-efficacycompared with their pre-workshop scores(p<0.05). Knowledge scores were alsosignificantly higher in both groups (p<0.01).Applied nutrition knowledge scores werehowever greater for the diet guide group(p<0.01).
Tailored: NoGroup Setting: NoFeedback: NoPsychological: NoPrimary MD: No
Excluded from meta-analysis as no relevantoutcome.
Of diabetic patients receiving drug informationsheets (arm 2), those who recalled receipt hadthe greatest improvement in follow-up testscores (4.53 to 6.16, p<0.001) but follow-up testscores were significantly higher (p<0.001) inboth intervention group (arm 2) and usual caregroup (arm 1).
Follow-up times: 2 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
174
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
InterventionSample Size
InterventionCharacteristics
Meta-Analysis Data* or OutcomesFollow-up Time(s)
1 Control (n/a)Counseling/therapy (Office visit)Education (Office visit)
Tailored: YesGroup Setting: NoFeedback: YesPsychological: NoPrimary MD: No
Insufficient statistics for meta-analysis.
Foot care education, behavioral contracts, andreinforcement (arm 2) resulted in 0.41 timesfewer serious foot lesions and more appropriatefoot care behavior (p=0.0001). Interventionsubjects (arm 2) were also more likely to havefoot examinations than were those in the usualcare group (arm 1) (68% vs. 28%, p<0.001).
Follow-up times: 12 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
176
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
InterventionSample Size
InterventionCharacteristics
Meta-Analysis Data* or OutcomesFollow-up Time(s)
1 Control (n/a)Education (One-on-one)Education (Reading material)Education (n/a)Feedback (n/a)
Excluded from meta-analysis as no usual careor comparable control group.
Patient education utilizing videotapes (arm 1)had significant weight loss at 7 monthscompared to education without videotapes(arms 2 and 3), but changes were not sustainedat 11 months. There were no significantchanges in HbA1c.
Follow-up times: 7 MO, 11 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
178
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: YesPsychological: YesPrimary MD: No
Follow-up time not in 3 - 12 months.
No differences in weight or glycosylatedhemoglobin were noted between interventiongroups (arms 2 and 3) and usual care group(arm 1) at 8- or 16-week follow-up.
Follow-up times: 8 WK
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
179
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Patients attending a behavior modification group(arm 2) had greater weight loss than those inindividual counseling (arm 1) at 12 weeksfollow-up (p<0.05) but had higher triglyceridelevels (p<0.10). Fasting serum glucose was notappreciably different between groups.
Excluded from meta-analysis as no usual careor comparable control group.
Self-management participants (arm 2) hadcontinued weight loss at 2, 3 and 6-monthfollow-up, compared to those receivingconventional treatment (arm 1), who, onaverage, gained weight. Fasting blood glucosesignificantly decreased for both groups overtime but was not significantly different betweengroups. Systolic and diastolic blood pressuresincreased in both groups over time but less sofor self-management subjects. Satisfactionmeasures showed no differences.
Follow-up times: 1 MO, 2 MO, 3 MO, 6 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
180
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: NoPsychological: NoPrimary MD: No
Excluded from meta-analysis as no relevantoutcome.
Intervention subjects (arm 2) had no significantdifferences compared to usual care group (arm1) with respect to diabetes-relatedhospitalizations over a 12-month period.Similarly, length of hospitalization, emergencyroom visits, and physician visits were nodifferent between groups despite significantgains for intervention subjects in self-careknowledge and skills in individual subject areasas well as in aggregate (p<0.001).
Follow-up times: 6 MO, 1 YR
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
181
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
At 6 months, HbA1c levels decreased 1.3% forintervention subjects (arm 2) as compared to0.22% for usual care group (arm 1). Interventionlevels persisted at 12 months but control levelshad fallen to similar levels by then as well. Self-care practices, self-efficacy, and satisfactionwith diabetes care were also greater forintervention subjects compared with usual caregroup.
Follow-up times: 6 MO, 18 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
182
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Patients in all groups were consulted by anutritionist. Three intervention groups (arms 2,3, and 4) received dietary plans that differed inrecommendations for fiber and oat bran intake.These groups demonstrated decreased bodyweight at 6-week follow-up for the oat brangroup (arm 4) compared to controls (arm 1)(p<0.05). Glycosylated hemoglobin decreasedin all 3 dietary groups but only in the increasedfiber group (arm 3) was this differencestatistically significant compared with controls(p<0.05).
Follow-up times: 2 WK, 6 WK
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
183
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: NoFeedback: YesPsychological: NoPrimary MD: No
Follow-up time not in 3 - 12 months.
Significant improvements for patients receivingeducation (arm 2) were noted in HbA1c, fastingplasma glucose, weight and blood pressure, butgreatest improvements were noted in the groupreceiving both patient and physician education(arm 4).
Follow-up times: 26 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
184
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Thirty minutes of professional instruction forself-monitoring blood glucose with Chem-stripbG (arm 2) compared with reading packageinstructions and practice (arm 1) resulted only ina lower percent error in blood glucoseestimation (p<0.02).
Subjects receiving diabetes counseling (arm 2)had significant reductions in post-prandial bloodglucose compared with usual care group (arm1) (p=0.009) at 6-month evaluation.
Follow-up times: 6 MO
1 Control (n/a)Counseling/therapy (One-on-one)Education (Group meeting)
Excluded from meta-analysis as no usual careor comparable control group.
Patients randomized to a behavior modificationgroup (arm 2) lost more weight than nutritioneducation (arm 3) or standard care (arm 1)groups during a 4-month treatment period(p<0.01). However, 16 months later, differencesin weight loss across these 3 groups were notsignificant.
Follow-up times: 4 MO, 10 MO, 16 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
188
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Two groups of randomized patients: a standardbehavioral weight control program (arm 1) and aweight control program that included self-monitoring of blood glucose and instruction inthe relationship between weight and glucoselevels (arm 2), both demonstrated significantweight loss (mean of 6.3 +/- 4.0 kg) at 12 weeksbut with no difference between groups.Significance was not maintained at one year.Nor were there any differences between groupsfor glycosylated hemoglobin, fasting bloodglucose, total cholesterol, blood pressure,medication use, or depression scores.
Follow-up times: 12 WK, 62 WK
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
189
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
InterventionSample Size
InterventionCharacteristics
Meta-Analysis Data* or OutcomesFollow-up Time(s)
1 Control (n/a)Dietary monitoring (Group meeting)Education (Group meeting)Financial incentives (Group meeting)Follow up (Group meeting)Goal setting (Group meeting)Material incentive (Group meeting)Practice self care skills (Group meeting)
Excluded from meta-analysis as no usual careor comparable control group.
After a 16-week treatment program, both self-regulation group (arm 2) and monitoring onlygroup (arm 1) significantly improved inbiochemical and weight measures but with nodifferences between arms. Though weight losswas significant for both arms at one-year follow-up, lack of difference between arms remained.HgbA1c values were unchanged in both armscompared to pretreatment values.
Follow-up times: 16 WK, 1 YR
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
190
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Significant decreases in HbA1c levels wereseen for individuals participating in computer-based interactive teaching programs withfeedback (arm 2) compared with usual caregroup (arm 1) (p<0.05). Knowledge increased inthese groups as well.
Follow-up times: 5 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
191
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Hospitalized patients receiving a comprehensiveinpatient diabetes education program (arm 2)had better compliance compared with controlgroup (arm 1) at 4-month follow-up with regardto self-care behaviors including exercise, insulinadministration and diet, however only exercisereached statistical significance (p=0.05). Bloodglucose was also lower (p=0.10) as was thenumber of emergency room visits (20 forcontrols versus 2 in experimental program,p=0.005).
Follow-up times: 1 MO, 4 MO
Evidence Table 1: Diabetes (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
192
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: NoGroup Setting: YesFeedback: NoPsychological: NoPrimary MD: No
Insufficient statistics for meta-analysis.
Though knowledge of arthritis and use ofexercise increased for both intervention groupscompared with no intervention, delivery byprofessional compared with layperson resultedin no differences with respect to pain,depression, physical function, social support ornon-exercise behaviors.
Excluded from meta-analysis as no usual careor comparable control group.
Patients receiving spouse-assisted coping skillstraining (arm 2) had lower levels of pain andpsychological disability and higher self-efficacyand more frequent use of pain-coping strategiesafter 10 weeks of treatment than did thosereceiving the cognitive-behavioral intervention(arm 1). Subjects in the pain-coping skillstraining without spouse assistance (arm 3) hadhigher self-efficacy, coping, and maritaladjustment and lower pain and psychological
Evidence Table 2: Osteoarthritis (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
196
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: NoPsychological: YesPrimary MD: No
Keefe F J, 1990b(#908)
Osteoarthritis (OA only)
RCT
Jadad Score: 1
Diagnostic criteria:X-ray and MD
Comorbidities:Obesity
3 Education (Group meeting)
n Entered: 36n Analyzed: 35
Tailored: NoGroup Setting: YesFeedback: NoPsychological: NoPrimary MD: No
Duplicate population Keefe F J, 1990b
Patients who received pain coping skills training(arm 2) had significantly lower levels of pain(p<0.01) and psychological disability (p<0.001)than those who received arthritis education (arm3) or usual care (arm 1). Physical disability wasno different between groups after treatment.
Follow-up times: 10 WK
Evidence Table 2: Osteoarthritis (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
198
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: NoPsychological: NoPrimary MD: No
Insufficient statistics for meta-analysis.
Subjects receiving relaxation procedures (arm4) had significantly less pain than thosereceiving other interventions or those in thecontrol group (p<0.05). No differences werenoted with respect to stiffness, mobility,medication taking behavior, or knowledge.
Follow-up times: 2 WK
Evidence Table 2: Osteoarthritis (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
199
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
2 Advocacy training (Telephone)Clinical reviews w/patient (Telephone)Reminders (Telephone)
n Entered: 109n Analyzed: 95
Tailored: YesGroup Setting: NoFeedback: YesPsychological: YesPrimary MD: No
3 Advocacy training (Office visit)Clinical reviews w/patient (Office visit)
n Entered: 109n Analyzed: 99
Tailored: YesGroup Setting: NoFeedback: YesPsychological: YesPrimary MD: No
Weinberger M, 1989(#430)
Osteoarthritis (OA only)
RCT
Jadad Score: 2
Diagnostic criteria:X-ray and MD
Comorbidities:n/a
4 Advocacy training (Office visit)Advocacy training (Telephone)Clinical reviews w/patient (Office visit)Clinical reviews w/patient (Telephone)Reminders (Telephone)
n Entered: 109n Analyzed: 97
Tailored: YesGroup Setting: NoFeedback: YesPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
Education delivered by telephone (arms 2 and4) compared with no telephone (arms 1 and 3)resulted in improved physical health andreduced pain (p=0.02) with trends suggestingimproved psychological health (p=0.10).
Follow-up times: 11 MO
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
202
Evidence Table 3: Post-Myocardial Infarction Care
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no relevantoutcome.
The average increase in functional capacity(i.e., peak treadmill workload on METS)between 3 and 26 weeks was significantlygreater (p<0.05) in training groups (arms2,3,4,5, and 6) than in the usual care group(arm 1) (1.8 vs. 1.2 METs, respectively).
Follow-up times: 3 WK, 11 WK, 26 WK
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
203
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
InterventionSample Size
InterventionCharacteristics
Meta-Analysis Data* or OutcomesFollow-up Time(s)
3 Counseling/therapy (One-on-one)Exercise diary (Self-delivery)Exercise monitoring (Telephone)Exercise program (One-on-one)Exercise program (Reading material)Exercise testing (n/a)Follow up (Telephone)
n Entered: 33n Analyzed: n/a
Tailored: YesGroup Setting: NoFeedback: YesPsychological: YesPrimary MD: No
4 Counseling/therapy (One-on-one)Exercise diary (Self-delivery)Exercise monitoring (Telephone)Exercise program (One-on-one)Exercise program (Reading material)Exercise testing (n/a)Follow up (Telephone)
n Entered: 33n Analyzed: n/a
Tailored: YesGroup Setting: NoFeedback: YesPsychological: YesPrimary MD: No
Excluded from meta-analysis as no usual careor comparable control group.
At 12 months, 4.1% had died in the interventionarm (arm 2), compared to 3.4% in the controlgroup (arm 1). LDL and total cholesteroldecreased more in the intervention arm(p <0.001). Smoking cessation at 12 monthsincreased significantly for the case managementarm versus usual care (70% vs. 53%, p=0.03).Functional capacity was higher in theintervention arm at 6 months 9.3 METS vs. 8.4METS. The% consuming a low fat dietincreased from 31% to 88% at 90 days in theintervention arm but was similar to usual carearm.
Follow-up times: 3 MO, 6 MO, 12 MO
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
205
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Nurse-delivered stress monitoring and stressreduction interventions resulted in lower stresslevels and fewer cardiac deaths (70% decrease)for intervention patients (arm 2) compared withusual care group (arm 1) but not reinfarctionrates. Differences between groups with respectto SES may be responsible for thesedifferences.
Follow-up times: 1 YR
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
206
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Subjects receiving home-based nursinginterventions aimed at reducing stress (arm 2)had significantly fewer MI recurrences thanusual care subjects (arm 1) over a 4-year follow-up period (p=0.04). The difference in mortalitywas maximal at 18 months post-MI, but duringthe remaining years mortality between groupswas equivalent. No difference in hospitalizationreadmission rates was noted.
Subjects receiving interventions of bothcardiologic and behavioral counseling (arms 2and 3) had lower 1-yr rates of reinfarction(p<0.01) and death (p<0.05) than usual caresubjects (arm 1). Behavioral counseling (arm 3)resulted in fewer reinfarctions (1.1% versus3.3% p<0.05) than cardiologic counseling alone(arm 2).
Follow-up times: 1 YR
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
207
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
Patients receiving Type A behavioral counseling(arm 2) had a 7.2% 3-year cumulative cardiacrecurrence rate compared with 13% forindividuals receiving only cardiologic counseling(arm 1) (p<0.005). Three-year survival withoutcardiac recurrence was also higher for thebehavioral counseling group (p<0.01) but nodifferences were noted for arrhythmias orhypertension.
Intervention subjects (arm 2) had 2.5 fewerhospital days (p<0.05), less observed weaknessand depression (p<0.05), decreased anxiety(p<0.001), and fewer supraventriculararrhythmias (p<0.05) compared with usual carepatients (arm 1). No differences in chest painoccurrence were noted.
Tailored: YesGroup Setting: NoFeedback: NoPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
Anxiety and general emotional disturbancescores for intervention subjects (arm 2) werehalf that of controls (arm 1) at 1-year follow-up.In the first 6 months of study, 18 controlcompared with 6 intervention patients hadhospital admissions (p=0.02).
Follow-up times: 6 WK, 6 MO, 12 MO
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
210
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no relevantoutcome.
Though functional capacity improved in patientsrandomized to either home (arms 3 and 4) orgroup (arms 5 and 6) exercise trainingcompared with controls (arms 1 and 2), nodifferences were seen between home and grouptraining. Frequency of exercise induced anginaor ischemic ST-segment depression was nodifferent between groups when measured at 26weeks.
Follow-up times: 3 WK, 11 WK, 26 WK
4 Counseling/therapy (One-on-one)Exercise diary (Self-delivery)Exercise monitoring (Telephone)Exercise program (One-on-one)Exercise program (Reading material)Exercise testing (n/a)Follow up (Telephone)
Both intervention groups (arms 2 and 3) hadimproved psychological measures related toanxiety, distress, and Type A behavior,compared with the usual care group (arm 1)(p<0.05). The counseling group (arm 3)demonstrated sustained significant reductions inalcohol and tobacco consumption at 12-monthfollow-up. A higher proportion of counselingsubjects reported returning to work by 12months and a trend towards less chest pain andrelated hospital admissions was also seen.
Follow-up times: 3 MO, 6 MO, 12 MO
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
212
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: YesPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
Subjects attending a behavioral group (arm 3)had statistically significantly less anxiety anddepression over 12-month follow-up than theusual care subjects (arm 1) (p<0.05). Type Abehavior was also reduced to a greater degreethan usual care or education interventionsubjects (p<0.01). Smoking decreased in allgroups but relapse rate for behavioral groupwas almost half that of the other 2 groups(p<0.05). The behavioral group also had fewerphysical symptoms and greater exercisecapacity (p<0.05).
Same study population as Sivarajan, et al.,1983. Using the Sickness Impact Profile surveyinstrument, researchers found improvedphysical and psychosocial function for thosereceiving an exercise program coupled withcounseling about cardiac risk factors andemotional adjustment after myocardial infarction(arm 3). Differences between groups exceededany changes noted for those receiving anexercise-only intervention and were significantat a .01 to .05 level dependent upon specificmeasured categories.
Follow-up times: 3 MO, 6 MO
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
214
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Chest pain frequency and depression scoreswere significantly lower for intervention subjects(arm 2) at 1-month follow-up but no differencesbetween intervention and usual care (arm 1)subjects were noted at 6-months.
Excluded from meta-analysis as no usual careor comparable control group.
Behavioral counseling (arm 2) targeted to “TypeA” life style resulted in greater reductions inType A behavior compared with standardcounseling (arm 1). Cardiovascular recurrencerates were no different between counselinggroups but behavioral counseling subjects hadlower 2-year cardiovascular recurrences thancontrols (2.76 versus 6.00p<0.05) Total cholesterol and blood pressurewere similar between groups.
Follow-up times: 2 YR
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
215
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
InterventionSample Size
InterventionCharacteristics
Meta-Analysis Data* or OutcomesFollow-up Time(s)
1 Control (n/a)Dietary monitoring (Office visit)Education (Reading material)
Same study population as Ott, et al., 1983.Though modest changes in diet were noted forintervention subjects (arms 2 and 3), nochanges occurred between groups with respectto weight or smoking.
Follow-up times: 3 MO, 6 MO
Evidence Table 3: Post-Myocardial Infarction Care (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
217
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: NoGroup Setting: YesFeedback: YesPsychological: YesPrimary MD: No
Excluded from meta-analysis as no relevantoutcome.
Subjective distress decreased in the stressmanagement group (arm 2) as compared to theusual care group (arm 1). This study lackedsignificant statistical power to detect potentiallymeaningful between-group differences.
Follow-up times: n/a
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
218
Evidence Table 4: Hypertension
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Excluded from meta-analysis as no usual careor comparable control group.
At 6-month follow up, significantly greaterdecreases were seen for both systolic BP anddiastolic BP in the relaxation arm (arm 2)compared with control arm (arm 1) (p<0.01,p<0.05, respectively).
Comorbidities:Heart disease, DM,arthritis, tobaccoabuse, and cancerand stroke
2 Education (Group meeting)Education (Instructional manuals)Education (Reading material)Follow up (One-on-one)Follow up (Telephone)Goal setting (One-on-one)
n Entered: 101n Analyzed: 95
Tailored: YesGroup Setting: YesFeedback: YesPsychological: NoPrimary MD: No
The intervention group (arm 2) had fewerdisability days and less self-reported functionaldecline but there were no differences based onphysical performance tests when compared withthe usual care group (arm 1). The number ofinpatient days was significantly less forintervention subjects (33 days versus 116 daysfor usual care group, p=0.049). Interventionsubjects also had greater physical activity(p=0.03) and less psychoactive medication use(p=0.04) than usual care group.
Evidence Table 4: Hypertension (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
226
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: YesFeedback: NoPsychological: YesPrimary MD: No
Insufficient statistics for meta-analysis.
Study subjects assigned to any of theexperimental groups had a 30% improvement inblood pressure control at 2 years and a 70%improvement at 5 years compared to 22% forthe usual care group with no difference inweight control or compliance in appointments.There was a 57% reduction in the 5-year all-cause mortality for intervention subjectscompared to those receiving usual care(p<0.05).
Follow-up times: 2 YR, 5 YR
Evidence Table 4: Hypertension (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
227
First AuthorYear(ID)
Condition (Type)Study DesignQualityPopulationCharacteristics Arm
Tailored: YesGroup Setting: NoFeedback: YesPsychological: YesPrimary MD: No
Diastolic BP (mmHg) at 24 weeks:Arm 1 = 94.5 (6.7)Arm 2 = 88.5 (6.7)Arm 3 = 90.0 (6.7)
Systolic BP (mmHg) at 24 weeks:Arm 1 = 138.0 (12.4)Arm 2 = 137.0 (12.4)Arm 3 = 137.8 (12.4)
Follow-up times: 8 WK, 6 MO
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
229
Evidence Table 5. Cost Articles
Articlenumber
Author/Year
Subjects (S), Follow-up period (F/U),Research design (D)and settings (ST)
Interventions Costs ofintervention
Effectiveness Health care costs orutilizations
C/E Ratings
Diabetes2270 Rettig et
al.,1986
S : 393 type1 and type2 diabetic patientsrecruited from amongdiabetic inpatients(mean = 52, 67%female)F/U: 6 and 12 monthsD: RCTST: Patient home
I: Needs assessmentand tailored individualinstruction at patienthome by a trained RNor LPN from homehealth nursingagencies.C: Usual care
Not reported, butinvolving 4-dayintensive course indiabetes self-care forparticipating nurses,and several homevisits (no more than12 for eachindividual).
At 6 months, interventionsubjects showedsignificantly greater self-careknowledge and skills thancontrol, although the actualdifferences in self-care skillswere probably too small tohave any practical meaning.No differences between thegroups were noted after 12mo of F/U.
At 6 and 12 months, nodifference was foundbetween control andintervention subjects interms of diabetes-relatedhospitalizations, length ofhospital stay, footproblems, emergencyroom and physician visits,and sick days.
Not cost-effective.
2159 Wood,1989
S: 93 hospitalizedpatients with type 1 or2 diabetes, age 20 to75 years old (mean 60,53% female).F/U: 1 mo, and 4months.D: RCTST: Hospital
I: Inpatient groupeducation programwhich stressed bothknowledge and self-help behaviors.C: Usual care
Not reported, buteach patient attendedtwo days of 2-houreducation program,with an averageattendance of four tosix patients. The 1st
session was taughtby a nurse educator,and the 2nd by aregistered dietitianand a communityhealth nurse.
Based on self-report. At 4month f/u, all respondentsreported a decline inperforming self-carebehaviors in comparisonwith the 1-month f/u.Compliance was lower forthe control group.Intervention group showedsignificantly bettercompliance than control inregards to exercise, diet,administering insulin, andbetter outcome measuresrelating to improvedmetabolic control andsignificant reduction in bloodsugar levels.
The intervention groupexperienced a significantlylower emergency roomvisitation rate (p <.005): At4 months, the 40 controlpatients reported 20 ERvisits, and the 53intervention patientsreported 2 ER visits. Thecontrol patients reported18 hospital readmission,and the interventionpatients reported 8hospital readmission.
Likely to becost-savings.
2589 deWeerdtet al.,1991
S : 558 insulin-treateddiabetic patients age18-65 years old (mean= 45)F/U : 6 monthsD: RCTST: 15 hospitals in
I: 1) Collaborativegroup education led byhealth-care worker(HCW), 2) Sameeducation led by fellowpatientsC: Usual care
Direct costs of theeducation program(including the costs ofemploying theeducators) andindirect costs (costsof the hours spent by
No significant effect ofeducation program onmetabolic control or qualityof life.
No significant effect ofeducation program oncosts of using healthservices (although theexperimental groupsshowed a trend to adecrease in the length of
Not cost-effective.Possiblereasonsinclude thequality of theeducation
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
230
Netherlands (5 forcontrol)
the participants inattending theeducation) togetherequal to US$100 perpatient (1990 dollar).Adding the cost ofdevelopingeducational materialsmake the per-patientcost to US$144 (1990dollar). No differencebetween I1 and I2.
hospitalization, but it wasnot significant). Almostequal changes in thenumber of visits to thephysician and GP werefound. No significantdifference in the dailyinsulin dosage andnumber of injections werefound between groups.Compared with the controlgroup, frequency of self-blood glucose monitoringincreased significantly inboth experimental groups.No significant effect ofeducation on the numberof sick days was found.
program, andthe lack ofsupportivechanges instandardtherapy andfollow-up of theeducationgiven.
2175 Kaplanet al.,1987
S : 76 volunteer adultswith type 2 diabetes(44 women), mean age= 55.F/U: 3, 6, 12, and 18months.D: RCTST: Community
I: Behavioral-basedgroup intervention.Each participant wasassigned to one of thethree 10-weekprograms: 1) diet, 2)exercise, 3) diet plusexercise.C: 10-week programsof group education.
Direct cost for dietand exercisecombined program isestimated to be$1000 (1986 dollar)per participant(including charges forhistory and physical,lab work, sessions,and medicalconsultations). This isnon-incremental cost.
70/76 completed follow-upstudy. At 18 months, thecombination diet-and-exercise group had achievedthe greatest reductions inglycosylated hemoglobinmeasures. In addition, thisgroup showed significantimprovements on a generalquality of life measure, equalto 0.092 incremental yearsof well-being for eachparticipant compared tocontrol.
N/A. Authorsreportedcost/utility =$10870/wellyear. However,cost is notcalculatedincrementally(if so, the C/Urate would bemorefavorable).
0749 Arsennau et al.,1994
S : 40 patients (mean= 59) attendingdiabetes educationprogramF/U: 2 and 5 monthsD: RCTST : Hospital
I : Individualizedlearning activitypackages (LAP)C: Classroominstruction
Instruction at thehospital costs $31 perhour (1995 dollar),The three LAPs weredeveloped to require3.5 hours ofinstructional time.Thus using LAPscould save individuals
At the 5-month f/u, the LAPgroup scored significantlyhigher on knowledgeassessment and decreasedpercent of ideal body weight.Patients who receivedclassroom instructionexhibited significantlydecreased glycosylated
Not studied. LAPs couldprovide acheapermeans of edu.,but lesseffectivenessin loweringblood glucoselevels than
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
231
$108.50 ininstructional fees.
hemoglobin levels. classroom edu.
2586 Campbell et al.,1996
S : 238 type 2 DMpatients, 80 years oryounger (mean = 58,51% female) withoutprevious formalinstruction in diabetescare.F/U: 3, 6, and 12monthsD: RCTST: Patients werereferred to a DiabetesEducation Service foreducation programs.Behavioral programwas conducted inpatient home.
I: Comparing relativeeffectiveness of thefollowing programs: 1)minimal instructionprogram, 2) educationprogram of individualvisits, 3) educationprogram incorporatinga group educationcourse, 4) behavioralprogram. (Note: 2) and3) are standard care.)
Not reported, butinvolving 1) two 1-hour sessions, 2) twoinitial sessions, and30 minuets monthlysession for 1 year, 3)at least two individualsessions and a 3-daysmall group educationcourse, as well astwo-hour groupfollow-ups at 3 & 9months, 4) 6 or moreindividual visits from anurse educator
Individual and groupeducation programs hadhigher attrition rates (40%)than the behavioral andminimal programs (10%). Nodifferent outcomes werefound between groups interms of physiologicalmeasures and BMI, exceptfor behavioral programproduced a greaterreduction in diastolic bloodpressure over 12 mos and agreater reduction in thecholesterol risk ratio over 3mos. The behavioralprogram patients reportedhigher satisfaction.
There were no differencesbetween groups overthree time periods inproportion of patientsconsulting anophthapmologist. Thebehavioral programpatients were more likelyto have visited a podiatristafter 6 months. Thegroups did not differ interms of a mean numberof visits they had made toa general practitioner , inhospital admissions, or inthe proportion who hadchanged the intensity oftheir blood pressuretreatment.
S : 206 diabeticpatients 40 years andolder (mean age = 62,62% female)F/U: 12 monthsD: RCTST: Outpatient clinics
I: Individualized,medical office-basedintervention focusedon dietary self-management, involvedtouch screencomputer-assistedassessment thatprovided feedback onkey barriers to dietaryself-management, goalsetting and problem-solving counseling.Follow-up componentsincluded phone callsand videotapeintervention relevant toeach participant.C: Usual care
From the perspectiveof a health careorganization, theincremental cost forthe delivery of theintervention totaled$14,755, or $137 perparticipant (1995dollars).
The intervention producedsignificantly greaterimprovement than usualcare on multiple measuresof change in dietarybehavior (e.g., covariateadjusted difference of 2.2%of calories from fat; p=0.023) and on serumcholesterol levels (covariateadjusted difference of15mg/dl; p = 0.002) at 12-month follow-up. Therewere also signicicantdifferences favoringintervention on patientsatisfaction (p < 0.02). Nosignificant improvement oneither HbA1c or on BMI.
Not studied. $7-$8 permg/dlreduction incholesterolcompare wellto estimates ofalternativeinterventionincludingcholesterolloweringmedications,which can costfrom $350 to$1400 perpatient year.
1668 Sadur et S : 185 patients of a I: Multidisciplinary Not reported, but may After the intervention, HbA1c Intervention group For patients
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
232
al.,1999
HMO aged 16-75(mean = 56, 43%female) and had eitherpoor glycemic controlor no HbA1c testperformed during theprevious yearF/U : 6 and 12 monthsafter randomizationD: RCTST: Outpatient clinic
outpatient diabetescare managementdelivered by a diabetesnurse educator, apsychologist, anutritionist, and apharmacist in clustervisit settings of 10-18patients/month for 6months.C: Usual care
not be more costlythan usual care since3 providers saw 12-18 patients for a 2-hour session monthly(somewhat highernumber of patientsthan these sameproviders would seein one-on-onesessions during thesame 2 hour), andmodestly reducedphysician visits.
levels declined significantlyin the intervention subjectscompared to controlsubjects. Several self-carepractices and severalmeasures of self-efficacyimproved significantly in theintervention group.Satisfaction with theprogram was high.Limitation: Failure to obtainfollow-up HbA1c levels andquestionnaires on 16% and25% of subjectsrespectively.
patients had somewhathigher ambulatory careutilization and moreintensive pharmaceuticalmanagement than controlsubjects during the 6-month intervention. Thisexcess utilization wasoffset by fewer hospitaladmissions after theintervention. Both hospitaland outpatient utilizationwere significantly lower forintervention subjects afterthe end of the program.
who had poordiabeticmanagement,providing thisintensivemanagementprogram maybe cost neutralin the shortterm (< 2years).
0828 Litzelman et al.,1993
S : 395 patients withtype 2 DM whounderwent the initialpatient riskassessment (352completed the study)(mean age = 60, 81%female, most subjectsare poorly educatedand indigent blackwomen)F/U : Completion ofintervention (12 monthfrom initialassessment)D: RCTST: Academicoutpatient clinic
I: Multifaceted,including 1) patienteducation andbehavioral contractabout foot-care, andalso reinforcementreminders, 2) healthcare system support ofidentifiers on patientfolders to promptproviders, 3) givenproviders practiceguidelines andinformational flowsheets on foot-relatedrisk factors foramputation.C: Usual care
The study materials,including folders, footdecals, postage,printing, andeducational materials,cost less than $5000.The major expense ofthe study was thesalary support for thenurse-clinicians whodid the assessmentsand for the researchassistant whoprocessed the charts.
Patients receiving theintervention were less likelythan control patients to haveserious foot lesions (oddsratio 0.41, p = 0.05) andother dermatologicalabnormalities. Also theywere more likely to reportappropriate self-foot-carebehaviors, to have footexaminations during officevisits (68% vs. 28%, p <0.001), and to receive foot-care education from healthcare providers (42% vs.18%, p < 0.001). Physiciansassigned to interventionpatients were more likelythan physicians assigned tocontrol patients to examinepatients’ feet.
At the end of theintervention, fouramputations had beendone in the control groupcompared with one in theintervention group.(Incidence rate is toosmall to test statisticalsignificance). Physiciansassigned to interventionpatients were more likelythan physicians assignedto control patients to referpatients to the podiatryclinic, but no difference inthe pattern of patientreferral to orthopedics andvascular surgery clinics.
S : 100 subjects witharthritis. 85 completedstudy. Mean age = 64.73% female. 73% had
I1 = An Arthritis Self-Management course(ASM) group taught bya male rheumatologist
The 12-hour coursetaught by 2 lay-leaders would costfrom $0.00
Professional-taught groupsdemonstrated greaterknowledge gain while lay-taught groups had greater
No significant difference innumber of visits tophysician at 4 monthsfollow-up between groups
Lay-taughtASM coursecould be aseffective as
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
233
OA.F/U: 4 monthsD: RCTST: Community sites
and a female physicaltherapist.I2 = An ASM coursegroup taught by 2female lay-leaders.C: No intervention.
(volunteer) to $200(1985 dollars). By onehealth professional,the course would costfrom $240 ($20/h) to$600 ($50/h). Thecosts of training andsupport for lay-leaders were notaccounted.
changes in relaxation thanthe other two groups. Thesubjects who received ASMcourse were more likely toexercise, and a tendencytoward less disability thancontrol subjects.
or change from baseline. professional-taught yetcheaper.However, bothfailed todemonstratereduction innumber ofphysicianvisits.
0835 Lorig etal.,1985
S : 190 subjects witharthritis. Mean age =67. 83% female. 77%had OA.F/U: 4 months RCTand 20 monthslongitudinal study.D: RCT + longitudinalstudyST: Community sites
I: An Arthritis Self-Management course(ASM) given in 6sessions by laypersons, based on astandardizededucational protocolemphasizing groupdiscussion, practice,the use of contractsand diaries to improvecompliance, andweekly feedback. Nosubsequentreinforcement. (129subjects)C: Delayedintervention for 4months. (61 subjects)
$15 to $20 perparticipant. (1983dollars).
At 4 months, experimentalsubjects significantlyexceeded control subjects inknowledge, recommendedbehaviors, and in lessenedpain. These changesremained significant at 20months.
At 4 months, there was atendency of decline invisits to physicians by theintervention group, but didnot reach statisticalsignificance at .05 level.The 20 monthslongitudinal study showedthe number of physicianvisits reduced frombaseline to 4-month f/u,and from 4 months to 8-months, and remainedabout the same from 8months to 20 months.These changes did notreach statisticalsigniciance.
S : 211 patients withknee OA from thegeneral medicine clnicof a municipal hospital(Of which 25 lost tof/u). Mean age = 63.85% female.F/U: 1 yearD: CT (NonrandomizedAttention-controlledclinical trial)
I : Self-care education:Individualizedinstruction and follow-up emphasizingnonpharmacologicmanagement of jointpainC: A standard publiceducation presentationand attention-controlling follow-up.
The cost ofdeliverying the self-care educationintervention to 105subjects was $6,163(in 1996 dollars), orequivalently, $58.70per patient.
See health care costs orutilizations.
The 94 subjects remainingin intervention groupmade 528 primary carevisits during the follow-upyear, while the 92controlled patients made616 visits. The averagesubject in interventiongroup generated $262 inclinic costs, comparedwith $322 for the average
For more than50% ofpatientsreceiving theintervention,the reducedoutpatientvisits and costsoffset theinterventioncosts. 80% of
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
234
ST: Outpatient clinic subject in control group.The frequencies and costsassociated with chargesfor drugs, radiography,and laboratory tests weresimilar between groups.
theinterventioncosts wasoffset withinone year dueto reducedoutpatientvisits.
Groessl&Cronan,2000
S : 363 members of aHMO, 60 years of ageand older with OA.Mean age = 70.F/U: 3 yearsD: RCTST: Community
I1: Social supportI2: EducationI3: A combination ofsocial support andeducationC: Usual care
$9450 for socialsupport group,$18675 for educationgroup, and $14175for combinationgroup, totaling$42300 (all in 1992dollars).
Feelings of helplessnessdecreased in theintervention groups but notin the control group. Allgroups showed increases inself-efficacy and overallhealth status.
Health care costsincreased less in theintervention groups than inthe control group. Basedon the HMO data, healthcare cost savings were$1,156/participant for yearone and two, and$1,279/participant for yearthree (1992 dollars).
S: 50 (24 female) adultaverage 51.1 years ofage with essentialhypertension recruitedvia mediaannouncements.Secondaryhypertension or withmean baseline bloodpressures greater than180 mm Hg systolic or120 mm Hg diastolicwere excluded.F/U: 6 weeksD: RCTST: Patient home
I: 1) individualizedrelaxation (IR), 2)group relaxation (GR),3) group relaxationplus contingencycontracting for homepractice (GRCC)C: Waiting list control
Measured bytherapist time
Measured by percentreductions in systolic anddiastolic blood pressure, andby eliciting home relaxationpractice
Not measured. GR wassignificantlymore costeffective thanIR for systolic,whereas bothGR and GRCCwere morecost effectivethan IR fordiastolic bloodpressure. Foramount ofrelaxationpractice, GR >GRCC > IR.
Post Myocardial Infarction Care2669 DeBusk
et al.,1985
S: 198 men 70 yearsor younger, had hadclinicallyuncomplicated AMI,
I1A: Medically directedat-home rehabilitationtraining for 23 weeksI1B: Medically directed
Three months of at-home rehabilitationwas estimated to beapproximately $328
Compared to the grouprehabilitation, medicallydirected at-homerehabilitation had about
Not studied. Medicallydirected at-homerehabilitation
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
235
mean age 52 ± 9years.F/U: 26 weeksD: RCTST: Patient home orcommunity gymnasium
at-home training for 8weeksI2A: Supervised grouptraining in agymnasium for 23weeksI2A: Supervised grouptraining in agymnasium for 8weeksI3: Exercise testingwithout subsequentexercise trainingC: Neither testing nortraining
per patient (1982 or1983 dollar). Thegroup rehabilitationprogram wasapproximately $720.
equally high adherence toindividually prescribedexercise, increase infunctional capacity, and lownonfatal reinfarction anddropout rates. Compared tothe no-training and controlgroups, the training groupswere significantly greater infunctional capacity, but notdifferent in cardiac events.
has thepotential todecrease thecost ofrehabilitatinglow-risksurvivors ofAMI.
0827 Lewin etal.,1992
S: 176 male andfemale patients with anAMI and age less than80 years (mean age =55.8 ± 10.6 years)F/U: 1 yearD: RCTST: Patient home
I: A comprehensiveself-help rehabilitationprogramme based ona heart manualSpouses were givenmaterials to supportand encouragecompliance bypatients. Includedfollow-up andfeedback.C: Standard care plusa placebo package ofinformation andinformal counseling.
The authorsestimated the cost oftreatment per patientto be £30 - £50 (1990dollar).
Psychological adjustmentwas better in therehabilitation group at 1year. The improvement wasgreatest among patientswho were clinically anxiousor depressed at dischargefrom hospital.
The two groupssignificantly differed in thenumber of GPconsultations at sixmonths and after thesecond six months; thecontrol group made amean of 1.8 more visitsthan did the rehabilitationgroup in the first 6months, and a mean of0.9 more visits in thesubsequent 6 months. Inaddition, significantly morecontrol patients thanrehabilitation grouppatients were admitted tohospital in the first 6months (18 vs. 6) but notat 12 months (18 vs. 9).Significantly fewerrehabilitation grouppatients were readmittedto hospital in the first 6months (8% vs. 24%).
Based onphysician self-report data foruse of healthservices.Indicative cost-saving.
Non-disease-specific Programs
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
236
1175 Leveilleet al.,1998
S: 201 chronically illseniors aged 70 andolder (mean age =77.1 years) with heartdisease, high bloodpressure, arthritis,cancer, stroke, ordiabetes. More % offemale in interventionthan in control (63.4%vs. 48.0%)F/U: 1 yearD: RCTST: A large seniorcenter, in collaborationwith primary careproviders of MCOs.
I: A geriatric nursepractitioner (GNP) ledmulti-componentprogram including riskfactor and healthassessment, feedbackto PCPs, follow-upvisits and phonecontacts, physicalactivity for disabilityprevention, andindividual counselingabout disease self-management as wellas group classes.C: Access to all seniorcenter activities, but noGNP.
The authorsestimated theprogram cost(primarily the salariesfor the GNP and thesocial worker) to beapproximately $300(1997 dollar) annuallyper participant.
The intervention groupshowed less decline infunction, as measured bydisability days and lowerscores on the HealthAssessment Questionnaire.However, the measures bySF-36 and a battery ofphysical performance testsdid not show difference byintervention. Theintervention led tosignificantly higher levels ofphysical activity and seniorcenter participation.
The number ofhospitalized participantsincreased by 69% (from13 to 22) among thecontrols and decreased by38% (from 21 to 13) inthe intervention group (p =.083). The total number ofinpatient hospital daysduring the study yeardecreased by 72% in theintervention group butincreased by 21% in thecontrol group(p = .049). The 83 lesshospital days in theintervention group yieldeda savings ofapproximately $1200 perparticipant. Outpatientvisits did not change in theintervention group butslightly increased in thecontrol group. There weretwo less ER visits in theintervention group but 8less ER visits in thecontrol group.
Indicative costsaving, due toless hospitaluse.
1510 Coleman etal./1999
S: 169 patients aged65 and older (mean =77) with the highestrisk for beinghospitalized orexperiencing functionaldeclineF/U: 2 yearsD: RCTST: Nine primary carephysician offices in alarge staff-model HMO
I: Chronic Care Clinicsattempted toreorganize the deliveryof primary careservices to better meetthe needs of olderpersons with chronicillness, includingdisease managementplanning, medicationreview, patient self-management/supportgroup)
Not available. After 24 months, nosignificant improvements infrequency of incontinence,proportion with falls,depression scores, physicalfunction scores, orprescriptions of high riskmedications weredemonstrated. A higherproportion of interventionpatients rated the overallquality of their medical careas excellent compared with
At baseline, interventionpatients were more likelyto be hospitalized. Duringthe 24-month follow-up,costs of medical careincluding frequency ofhospitalization, hospitaldays, emergency andambulatory visit, and totalcosts of care were notsignificantly differentbetween intervention andcontrol groups.
Insufficientinformation(implied notcost saving).
Evidence Table 5: Cost Articles (con't)
N/A = Not Available or Not ApplicableNOS = Not Otherwise Specified* Unless otherwise specified, Mean (Standard Deviation) reported.
237
C: Usual care control patients (40% vs.25%, p = 0.1)
0608 Lorig etal.,1999
S: 952 patients 40years and older (mean= 65) with heartdisease, lung disease,stroke, or arthritis.F/U: 6 monthsD: RCTST: Community-basedsites (churches, seniorand communitycenters, publiclibraries, & health carefacilities)
I: Subjects receivedthe Chronic DiseaseSelf-ManagementProgram (CDSMP), acommunity-basedpatient self-managementeducation course. Thecontent andmethodology of theCDSMP were basedon needsassessments. Theprocess of teachingthe course is based onSelf-Efficacy Theory.The course was taughtby a pair of trained,volunteer lay leaders.C: Waiting list control
The authorsestimated theprogram cost to beapproximately $70(1998 dollar) perinterventionparticipant. Thisincludes $26 fortraining leaders, $14for volunteer leaderstipend, $15 forcourse materials, and$15 administrativecosts. This analysisdoes not take intoaccount the cost ofspace or indirectcosts.
At 6 months, treatmentsubjects demonstratedimprovements in weeklyminutes of exercise,frequency of cognitivesymptom management,communication withphysicians, self-reportedhealth, health distress,fatigue, disability, andsocial/role activitieslimitations, compared withcontrol subjects. Programeffects were similar acrossall four diagnosticsubgroups.
Based on patient self-report, the treatmentgroup reduced theirphysician visits slightlymore, but not significantly,than did the control group.However, the decrease inthe number ofhospitalizations and in thelength of hospital stayswere significant at p <.05.Assuming a cost of $1000per day of hospitalization,the 6-month health carecosts for each controlparticipant in this studywere $820 greater than foreach treatment subjects.