Long-term impact of home telehealth service on preventable hospitalization use Huanguang “Charlie” Jia, PhD Research Health Scientist VA RORC REAP North.

Long-term impact of Long-term impact of home telehealth service on home telehealth service on preventable hospitalization preventable hospitalization

useuseHuanguang “Charlie” Jia, PhDHuanguang “Charlie” Jia, PhD

Research Health ScientistResearch Health ScientistVA RORC REAP VA RORC REAP

North Florida/South Georgia VHSNorth Florida/South Georgia VHSGainesville, FloridaGainesville, Florida

Co-authorsCo-authors

Ho-Chih Chuang, MSHo-Chih Chuang, MS Samuel S. Wu, PhDSamuel S. Wu, PhD Xinping Wang, PhDXinping Wang, PhD Brad N. Doebbeling, MDBrad N. Doebbeling, MD Neale R. Chumbler, PhD Neale R. Chumbler, PhD

AcknowledgementAcknowledgement

This work was funded by the Community This work was funded by the Community Care Coordination Service at VA VISN 8 Care Coordination Service at VA VISN 8 through the Rehabilitation Outcomes through the Rehabilitation Outcomes Research Center (RORC REAP) at N. Florida/S. Research Center (RORC REAP) at N. Florida/S. Georgia VHS, Gainesville, FL.Georgia VHS, Gainesville, FL.

The views expressed in this report are those The views expressed in this report are those of the authors and do not necessarily of the authors and do not necessarily represent the views of Department of represent the views of Department of Veterans Affairs. Veterans Affairs.

Background: ACSC & Background: ACSC & Preventable Preventable HospitalizationHospitalization

Hospitalizations for ACSCs may be Hospitalizations for ACSCs may be prevented if timely and appropriate prevented if timely and appropriate ambulatory care were accessible.ambulatory care were accessible.

Barriers to accessibility include Barriers to accessibility include provider unavailability, costs, health provider unavailability, costs, health insurance absence.insurance absence.

Improved access at community level Improved access at community level would lower ACSC hospitalization.would lower ACSC hospitalization.

References:1) Weissman JS, et al. Rates of avoidable hospitalization by insurance status in Massachusetts and Maryland. JAMA. 1992;268:2388-23942) Bindman AB, et al. Preventable hospitalizations and access to health care. JAMA. 1995;274:305-3113) Culler SD, et al . Factors related to potentially preventable hospitalizations among the elderly. Med Care. 1998;36:804-8174) Friedman B, Basu J. The rate and cost of hospital readmissions for preventable conditions. Med Care Res Rev. 2004;61:225-2405) Basu J, et al. Primary care, HMO enrollment, and hospitalization for ACSCs. Med Care. 2002;40:1260-1269

Background: Home Background: Home TelehealthTelehealth

Application of modern Application of modern telecommunications.telecommunications.

Link patients to out-of-home sources of Link patients to out-of-home sources of care information, education, or service.care information, education, or service.

Medical benefit: early detect problems, Medical benefit: early detect problems, frequently monitor conditions, increase frequently monitor conditions, increase access, improve care plan compliance. access, improve care plan compliance.

Home telehealth reduces inpatient & ER Home telehealth reduces inpatient & ER use within short-term.use within short-term.

References:1) Koch S. Home telehealth--current state and future trends. Int J Med Inform. 2006;75:565-5762) Hailey D, et al. Systematic review of evidence for the benefits of telemedicine. J Telemed Telecare. 2002;8 (Supplement 1):1-303) Barnett TE, et al. The effectiveness of a care coordination home telehealth program.. Am J Manag Care. 2006;12:467-4744) Chumbler NR, et al. Evaluation of a home-telehealth program for veterans with diabetes. Eval Health Prof. 2005;28:464-478

ObjectiveObjective

To test 4-year effect of a VA To test 4-year effect of a VA patient-centered, care patient-centered, care coordination/home telehealth coordination/home telehealth (CCHT) program on potentially (CCHT) program on potentially preventable hospitalization use preventable hospitalization use by veteran patients diagnosed by veteran patients diagnosed with diabetes mellitus. with diabetes mellitus.

Study Design

Retrospective, matched treatment and Retrospective, matched treatment and control study design. control study design.

Treatment group (n=387): DM patients, Treatment group (n=387): DM patients, enrolled in the CCHT program at 4 VAMCs.enrolled in the CCHT program at 4 VAMCs.

Control group (n=387): DM patients in the Control group (n=387): DM patients in the 4 VAMCs matched by a propensity score. 4 VAMCs matched by a propensity score.

References:1) Barnett TE, et al. The effectiveness of a care coordination home telehealth program for veterans with diabetes mellitus: A 2-year follow-

up. Am J Manag Care. 2006;12:467-4742) D'Agostino RB, Jr. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group.

Stat Med. 1998;17:2265-2281

VA CCHT ProgramVA CCHT Program

Transition from hospital-based care to Transition from hospital-based care to patient-centered and ambulatory care.patient-centered and ambulatory care.

Care coordination by nurse practitioner.Care coordination by nurse practitioner. Disease monitoring using supportive home Disease monitoring using supportive home

telemonitoring technology. telemonitoring technology. Each enrollee has a messaging device Each enrollee has a messaging device

installed at home using basic land-line installed at home using basic land-line telephone service. telephone service.

Daily basis: patients answer scripted Daily basis: patients answer scripted questions from the messaging device about questions from the messaging device about their diabetes symptoms and health status.their diabetes symptoms and health status.

Care coordinators monitor the patients’ daily Care coordinators monitor the patients’ daily updates from the devices.updates from the devices.

CCHT enrollment CCHT enrollment criteriacriteria

Diagnosed with DM.Diagnosed with DM. ≥≥1-time VA hospitalizations or 1-time VA hospitalizations or

≥≥1-time VA ER visits in 12 months 1-time VA ER visits in 12 months prior to enrollment.prior to enrollment.

Non-institutionalized.Non-institutionalized. A telephone land-line at home.A telephone land-line at home.

Dependent VariableDependent Variable

Semi-annual P.H. count by Semi-annual P.H. count by patient.patient.

AHRQ defined 12 ACSCs and ICD-AHRQ defined 12 ACSCs and ICD-9 codes applied.9 codes applied.

VA automated inpatient VA automated inpatient databases.databases.

References:1) AHRQ. Guide to prevention quality indicators: Hospital admission for ACSCs. March 12, 2007; Version 3.12) AHRQ. Prevention quality indicators: Technical specifications. March 12, 2007; Version 3.1

Independent & Independent & CovariatesCovariates

Treatment/CCHT enrollee: yes, no.Treatment/CCHT enrollee: yes, no. Baseline: age, gender, marital Baseline: age, gender, marital

status, race, VA care priority, and status, race, VA care priority, and study sites.study sites.

Pre-enrollment: 6-month Pre-enrollment: 6-month comorbidity score, 12-month comorbidity score, 12-month inpatient and outpatient use.inpatient and outpatient use.

Post-enrollment: 4-year survival Post-enrollment: 4-year survival time in days.time in days.

Statistical AnalysisStatistical Analysis

Descriptive statistics.Descriptive statistics. Multicollinearity diagnostics.Multicollinearity diagnostics. A GLIMMIX to estimate the impact A GLIMMIX to estimate the impact

of the CCHT program on P.H. use of the CCHT program on P.H. use over a period of 4 years, adjusting over a period of 4 years, adjusting for patient characteristics and for patient characteristics and time.time.

Table 1.1. Baseline Table 1.1. Baseline characteristics characteristics (No sig. difference observed)(No sig. difference observed)

CharacteristicsCharacteristics

Study Study cohortcohort

(N=774)(N=774) Tx (n=387)Tx (n=387) Ctrl (n=387)Ctrl (n=387)

AgeAge 67.6 (10.1)67.6 (10.1) 68.0 (9.2)68.0 (9.2) 67.2 (10.9)67.2 (10.9)

MaleMale 98.3%98.3% 98.7%98.7% 97.9%97.9%

Being marriedBeing married 61.9%61.9% 64.3%64.3% 59.4%59.4%

WhiteWhite 39.4%39.4% 40.1%40.1% 38.8%38.8%

HispanicHispanic 49.2%49.2% 48.8%48.8% 49.6%49.6%

High VA priorityHigh VA priority 97.9%97.9% 98.2%98.2% 97.7%97.7%

Site ASite A 14.6%14.6% 15.3%15.3% 14.0%14.0%

Site BSite B 14.5%14.5% 15.0%15.0% 14.0%14.0%

Site CSite C 46.3%46.3% 46.1%46.1% 46.5%46.5%

Site DSite D 24.7%24.7% 23.8%23.8% 25.6%25.6%

Table 1.2. Pre- & Post-Table 1.2. Pre- & Post-baseline Characteristicsbaseline Characteristics

CharacteristicsCharacteristicsStudy cohortStudy cohort(N=774)(N=774) Tx (n=387)Tx (n=387) Ctrl (n=387)Ctrl (n=387)

Pre-enrollment:Pre-enrollment:

6-m comorb score6-m comorb score 0.2(0.5)0.2(0.5) 0.3 (0.6)0.3 (0.6) 0.2 (0.5)0.2 (0.5)

12-m inpt. care use12-m inpt. care use 0.8(1.3)0.8(1.3) 0.7 (1.2)0.7 (1.2) 0.8 (1.5)0.8 (1.5)

12-m outpt. visit12-m outpt. visit‡‡ 26.5(21.6)26.5(21.6) 30.3 (21.7)30.3 (21.7) 22.6 (20.8)22.6 (20.8)

4-year post-enrollment:4-year post-enrollment:

P.H. countsP.H. counts†† 0.8(1.6)0.8(1.6) 0.7 (1.3)0.7 (1.3) 1.0 (1.9)1.0 (1.9)

Crude death rateCrude death rate†† 22.9%22.9% 19.4%19.4% 26.4%26.4%

Survival daysSurvival days‡‡ 1314(330)1314(330) 1349(266)1349(266) 1278(380)1278(380)

†† p value <0.05; p value <0.05; ‡‡ p value <0.01; p value <0.01;

Table 2. Freq of 4-year Table 2. Freq of 4-year P.H. ACSCs occurrences by P.H. ACSCs occurrences by group group

P.H. conditions/ACSCsP.H. conditions/ACSCs Diabetes long-term complicationsDiabetes long-term complicationsL. extremity amput. in DM ptsL. extremity amput. in DM ptsDiabetes short-term complicationDiabetes short-term complicationDiabetes uncontrolledDiabetes uncontrolledBacterial pneumoniaBacterial pneumoniaAnginaAnginaCongestive heart failureCongestive heart failureUrinary infectionUrinary infectionC. obstructive pulmonary diseaseC. obstructive pulmonary diseaseDehydrationDehydrationHypertensionHypertensionAdult asthmaAdult asthma

TxTx424229297744

222288

84843333313111113311

CtrlCtrl12112155552828151534341919676731311414995522

Table 3. Results from a Table 3. Results from a GLIMMIX GLIMMIX (dependent var=P.H. (dependent var=P.H. count) count)

CharacteristicsCharacteristics Relative Risk (95% CI)Relative Risk (95% CI) P valueP value

Treatment: yes vs. noTreatment: yes vs. no 0.36 (0.21-0.61)0.36 (0.21-0.61) 0.00020.0002

TimeTime 0.88 (0.82-0.94)0.88 (0.82-0.94) 0.00020.0002

Treatment x timeTreatment x time 1.15 (1.04-1.27)1.15 (1.04-1.27) 0.00470.0047

AgeAge 1.02 (1.00-1.04)1.02 (1.00-1.04) 0.02170.0217

Male: yes vs. noMale: yes vs. no 1.35 (0.37-4.85)1.35 (0.37-4.85) 0.64890.6489

Married: yes vs. noMarried: yes vs. no 0.54 (0.38-0.77)0.54 (0.38-0.77) 0.00060.0006

White: yes vs. noWhite: yes vs. no 0.64 (0.38-1.08)0.64 (0.38-1.08) 0.09490.0949

VHA priority high: yes vs. VHA priority high: yes vs. nono 0.87 (0.26-2.95)0.87 (0.26-2.95) 0.82840.8284

6-m prior comorbid score6-m prior comorbid score 1.54 (1.13-2.10)1.54 (1.13-2.10) 0.00600.0060

12-m prior inpatient use12-m prior inpatient use 1.57 (1.40-1.76)1.57 (1.40-1.76) <.0001<.0001

12-m prior outpatient 12-m prior outpatient visitvisit 1.01 (1.01-1.02)1.01 (1.01-1.02) 0.00060.0006

Site A vs. CSite A vs. C 0.97 (0.48-1.94)0.97 (0.48-1.94) 0.92550.9255

Site B vs. CSite B vs. C 1.68 (0.97-2.89)1.68 (0.97-2.89) 0.06300.0630

Site D vs. CSite D vs. C 1.50 (0.83-2.73)1.50 (0.83-2.73) 0.18180.1818

Main ResultsMain Results

The linear mixed results The linear mixed results suggest that the CCHT suggest that the CCHT enrollees were less likely to enrollees were less likely to be admitted for a P.H. (RR be admitted for a P.H. (RR 0.36, p<0.05). 0.36, p<0.05).

The difference reduced as The difference reduced as time progressed during the 4-time progressed during the 4-year follow-up.year follow-up.

LimitationsLimitations

A single geographic region. A single geographic region. VA healthcare system enrollees.VA healthcare system enrollees. Patients with DM, a diagnosis Patients with DM, a diagnosis

associated with high rates of associated with high rates of morbidity, mortality, and resource morbidity, mortality, and resource use. use.

ConclusionsConclusions

The VA CCHT program for The VA CCHT program for diabetes patients reduced diabetes patients reduced preventable hospitalizations preventable hospitalizations overtime.overtime.

It may reduce healthcare cost.It may reduce healthcare cost.