A Home Telemonitoring Program Reduced Exacerbation and Healthcare Utilization Rates in COPD Patients with Frequent Exacerbations

Abstract

Rationale: As the impact of home telemonitoring on patients with chronic obstructive pulmonary disease (COPD) is not fully understood and reduction in healthcare utilization is not proven, we attempt to evaluate the effects of home telemonitoring on healthcare utilization in patients with COPD. Subjects and Methods: We conducted a retrospective cohort study using the Veterans Health Administration database of COPD patients enrolled in the Care Coordination Home Telehealth (CCHT) program. We evaluated the effects of monitoring through this program in patients with moderate to severe COPD and frequent exacerbations. Numbers of emergency department (ED) visits, urgent care (UC) visits, and hospitalizations were all evaluated before and after enrollment. The differences in average pre-enrollment and during-enrollment numbers of hospital admissions, ED/UC visits, and exacerbations were tested for significance among all patients enrolled in the program who had one or more exacerbations at pre-enrollment; results were expressed on a per-year basis. Results: Data were available on a total of 1,133 patients with COPD enrolled in the CCHT program between 2005 and 2009. Given the objectives of our study, we only included 369 patients who had at least one exacerbation per year in the year prior to enrollment. Of these, 71.5% had a reduction in numbers of ED visits and exacerbations requiring hospitalizations after enrollment in the program. The average number of hospital admissions, ED visits, and total exacerbations were all reduced (0.41±1.68, 0.15±1.65, and 0.56±2.3, respectively; all with p<0.01). The pre-enrollment number of exacerbations was the only factor observed to be significantly associated with the reduction in number of exacerbations. Conclusions: In patients with COPD and frequent exacerbations, enrollment in a home telemonitoring program may decrease healthcare utilization.

Introduction

Chronic obstructive pulmonary disease (COPD) is a progressive disease that causes physical impairment, debility, and reduced quality of life. By the year 2020, it is predicted that COPD will be the third leading cause of death and fifth most common source of disability.¹ The National Heart, Lung, and Blood Institute in 2005 gave an estimate of the total (direct plus indirect) annual cost of COPD to the United States of $38.8 billion dollars, signaling the magnitude of COPD burden on healthcare and economy.² Most of these costs come from hospitalizations and emergency department (ED) visits for treatment of acute exacerbations. Appropriate treatment of stable COPD by use of inhalers in outpatient settings and early recognition of exacerbation and self-treatment prevent exacerbations and reduce ED visits and hospitalizations.^3,4

Many interventions are suggested to reduce costs of COPD care by preventing exacerbations and reducing ED and urgent care (UC) visits in a chronic care model.^5,6 One intervention to enhance home care is home telemonitoring, which has been associated with improved glycemic control and reduction in ED visits and hospitalizations in a cohort of Veterans Administration (VA) patients with diabetes.⁷ There is no consistent evidence in the literature regarding the degree with which home telemonitoring reduces the utilization of healthcare resources, and in some studies savings from reduced hospitalizations are offset by the costs to run the program including the salary of healthcare professionals.^8,9 A study from The Netherlands that used a home telemonitoring device (the same device used in the present study) to monitor COPD patients and included a control group showed trends toward reducing healthcare utilization. The study was underpowered, had a short follow-up period before and after enrollment, and could not show a definite reduction in healthcare utilization.¹⁰ A study similar to ours from the Veterans Health Administration system including patients with congestive heart failure and diabetes as well as COPD showed no significant benefit in healthcare utilization in COPD patients; the study was also underpowered (only 17 patients with COPD were included), and the enrollment period was very short.¹¹ More recent systematic reviews of published studies could not conduct meta-analyses on the included studies because of the heterogeneity of patient populations and diverse outcome measures reported. Most studies included in those reviews were underpowered, had short follow-ups, and had significant design biases. A clear benefit could not be drawn from those reviews.^12,13

In planning for this study we performed an initial analysis on the first 100 enrollees in the Care Coordination Home Telehealth (CCHT) program. This sample was part of our initial study population and included patients with other chronic illnesses in addition to COPD. The analysis revealed a trend toward reduction (albeit not statistically significant) in ED and unscheduled office visits (mean reduction of 0.14 visit per year, or a relative reduction of 14%; p=0.18). However; among 22 subjects in this group with two or more exacerbations per year there was a significant reduction (from 3.6 to 2.0 visits per year, or a 46% decrease; p=0.003) in ED and UC visits. These results were presented as an abstract at the American College of Chest Physicians meeting in 2009.¹⁴ We planned a larger sample to increase study power and length of enrollment.

Home telemonitoring is typically defined as the use of an electronic device by patients to enter data pertinent to their health condition, which in turn is communicated to and evaluated remotely by a healthcare professional; appropriate actions are then taken to prevent further progression and possible hospitalization. Our study attempts to evaluate the impact of a home telemonitoring program in patients with COPD in terms of reduction in healthcare utilization and possibly improve outcome. Our hypothesis is that this intervention would decrease the use of acute care services by facilitating communication between patients and their healthcare providers, providing outpatient interventions, and encouraging patient self-treatment.

Subjects and Methods

TECHNOLOGY USED

A device already in use for monitoring other chronic health conditions called iCare Health Buddy^® (Heath Hero Network, Inc., Palo Alto, CA) was adopted for use in monitoring and management of COPD as part of the CCHT program in 2005 by Veterans Administration Healthcare. Our study involved patients covered by the Veterans Integrated Service Network-16 (VISN 16). More than 445,000 veterans were served per year in VISN 16. The 10 VISN 16 centers and 40 community-based outpatient clinics served veterans in Oklahoma, Arkansas, Louisiana, Mississippi, and parts of Texas.

The iCare Health Buddy is a phone-based interactive system that asks patients questions and then records and transmits responses to the monitoring office via a home phone line. The device is programmed by the coordinator to ask patients about their respiratory symptoms such as cough, shortness of breath, wheezing, sputum production and characteristics, and need to use inhalation medications more often. The device also asks and records responses on daily activities, fever, and appetite. Responses are then transmitted to the coordinators' office and are analyzed within 24 h. Weekend and holidays responses are analyzed on the next workday. Responses are stratified by computer software according to the risks; responses with possible exacerbations are flagged as priority. Flagged responses are evaluated first; if an immediate action is deemed necessary, the primary care provider is notified. If the primary provider is not available, the pulmonologist supervising the program is contacted. Subsequent interventions are at the discretion of the caring physician. In most cases it involves maximizing inhalation therapy, prescription of antibiotics and steroids, or referral to an UC or ED site. Follow-up appointments are arranged in some instances with the primary care or pulmonary clinic at the discretion of the caring physician. The data collected also allowed for monitoring of trends in patients' symptoms and a proactive approach in case of a noticeable decline.

All patients were given one-time training on how to use the device at the time of enrollment along with explanation of the natural history of COPD based on the national clinical guidelines (the American Thoracic Society Guidelines). Some simple and safe self-management skills were also taught at the time of enrollment: specifically, the patient was instructed when to maximize his or her inhalers and report worsening symptoms while awaiting a call from the coordinator and when to report to the ED. An optional instructive 5-min message was also offered every time the device is turned on.

Study Design

This is a retrospective cohort study to evaluate the effects of home telemonitoring on exacerbation rates and healthcare utilization. Data from the VA database involving areas covered by the VISN 16 database were collected. In total, 1,133 patients were enrolled in the CCHT program under a diagnosis of COPD. However, data from only 369 subjects with one or more exacerbations per year before enrollment were analyzed. We limited our study to these patients as our objective was to determine if the program made a significant decrease in exacerbations and healthcare utilization rates among these patients. The dataset includes demographics, hospital admissions, and ED and UC visits with a primary ICD-9 diagnosis associated with a COPD exacerbation. Data collection started 1 year prior to enrollment through the end of 2009.

Study Participants

To participate in the program a patient must have a landline telephone and be able to use the device (free of visual, motor, and/or cognitive impairments). To be eligible for the program patients must have advanced COPD and high healthcare utilization as determined by their pulmonologists or primary care providers. Patients with frequent exacerbations (more than one per year) and patients with other chronic illnesses such as diabetes were targets for enrollment in this program. If a primary provider thought that a patient would benefit from the program, a consult for the CCHT program would be placed by the provider, and a nurse designated for the program would enroll patients, give instructions on how to use the device, monitor the responses on a daily basis, and communicate with the primary provider directly by phone or via the VA computerized patient records system.

Variables, Data Sources, Measurements, and Statistical Analysis

Among 1,133 patients enrolled in the CCHT program at the time of data collection, we analyzed data for 369 patients who had one or more exacerbations at pre-enrollment. All admissions and ED/UC visits were matched with electronic records and the primary ICD-9 diagnosis to ensure the correct diagnosis. Although patients enrolled in this program could have other chronic illnesses, data related to COPD alone were used in this study.

Demographic data included age at enrollment, race, sex, presence of other chronic diseases (diabetes, hypertension, or congestive heart failure), duration of enrollment (days), and smoking status. Some patients had missing race information. Exacerbation data included COPD-related (1) hospital admissions, (2) ED/UC visits, and (3) combined hospital admissions and ED visits. If the ED or UC visit occurred within 1 week of an admission it was counted as one exacerbation. Each patient had exacerbation data for pre-enrollment (the year before enrollment) and during enrollment. As a patient can be enrolled for less or more than a year in the program, the number of exacerbations during enrollment for each patient was expressed on a per-year basis by dividing enrollment duration by 365 in order to make number of exacerbations during enrollment comparable to the year before enrollment. For example, if a patient had one exacerbation per year in the year prior to enrollment, then had one exacerbation during enrollment, and was enrolled in the program for 500 days, then the exacerbation rate would be weighted as follow: 1/(500/365)=0.73 exacerbation per year, and the difference would be 1 – 0.73=0.27, which represents a positive decrease in exacerbations.

Differences between pre-enrollment and during-enrollment numbers of hospital admissions, ER visits, and exacerbations (sum of hospital admissions and ER visits) were analyzed. As the differences were non-normally distributed, nonparametric methods were used. The Wilcoxon signed rank test was used to test null hypotheses of no differences in the average numbers of hospital admissions, ER visits, and exacerbations (hospital admissions plus ER visits) the year before and during enrollment in the CCHT program. Each difference was calculated as a decrease from pre-enrollment, that is, number before enrollment minus number during enrollment. If the decrease in exacerbation was significant, the Wilcoxon rank sum test was used to determine categorical factors (gender, race, and smoking) significantly associated with the exacerbation decrease, and the Spearman rank correlation coefficient was used to determine numeric factors (age and pre-enrollment number of exacerbations) significantly correlated with exacerbation decrease. The chi-squared test and logistic regression analysis were used to determine factors associated with whether or not there was a positive decrease in the number of exacerbations brought about by the program. SAS software (version 9.2; SAS Institute, Cary, NC) was used in our statistical analysis. The study was approved by the Institutional Review Board at the Overton Brooks VA Medical Center, Shreveport, LA, on April 30, 2009, and data were extracted from the VISN 16 Data Warehouse on October 20, 2009. Informed consent was waived.

Results

Statistics on patient characteristics and outcomes are shown in Table 1. The majority of patients were male and white. Mean decreases in numbers of hospital admissions, ED visits, and exacerbations from year before enrollment are all positive. The percentage of patients whose number of exacerbations decreased during enrollment in the program is 71.5%. We determined factors significantly associated with the magnitude of decrease in number of exacerbations and with whether or not the number of exacerbations decreased during enrollment in the CCHT program.

Table 1.

Statistics on Patient Characteristics and Outcomes (n=369)

CHARACTERISTIC/OUTCOME	NUMBER (%) OR MEAN	MEDIAN	IQR
Male	351 (95.1)
White race^a	302 (84.1)
Had other diseases (DM, HTN, or CHF)	117 (31.7)
Smoker	138 (37.4)
Decrease in number of exacerbations >0	264 (71.5)
Died	95 (25.8)
Age (years)	69.3	68.0	61–78
Length of enrollment (days)	491.6	309.9	160–778
Decrease in number of
Hospital admissions^b	0.41	0.73	0.0–1.0
ED visits	0.15	0.0	−0.26 to 1.00
Exacerbations	0.56	1.00	−0.08 to 1.68

Missing race information on 10 patients.

Number of hospital admissions before enrollment – number during enrollment.

CHF, congestive heart failure; DM, diabetes mellitus; ED, emergency department; HTN, hypertension; IQR, interquartile range (first to third quartile).

Comparisons (using the Wilcoxon signed rank test) on number of hospital admissions, ED visits, and exacerbations between year before and during enrollment in the program are shown in Table 2; significant decreases in hospital admissions, ED visits, and exacerbations were noted.

Table 2.

Results on Decreases in Number of Exacerbations Between Before and During Enrollment

DECREASE IN	MEAN±SD	MEDIAN	RANGE	P VALUE^a
Number of hospital admissions	0.41±1.68	0.73	−11.70 to 5.62	<0.01^b
Number of ED visits	0.15±1.65	0	−8.47 to 7.43	<0.01^b
Number of exacerbations	0.56±2.36	1.0	−11.25 to 9.43	<0.01^b

For testing the null hypothesis of zero decrease.

Significant at the 1% level.

ED, emergency department; SD, standard deviation.

The only factor observed to be significantly associated with the amount of decrease in number of exacerbations is pre-enrollment number of exacerbations (Spearman correlation coefficient=0.41, p<0.01). The greater the pre-enrollment number of exacerbations, the greater is the decrease in number of exacerbations during enrollment. Test results for the decreases in number of hospital admissions, ED visits, and exacerbations among patients with two or more exacerbations before enrollment are shown in Table 3. The mean decreases were larger than those among patient with one or more pre-enrollment exacerbations (shown in Table 2).

Table 3.

Test Results on Decreases in Number of Exacerbations Among Patients with Two or More Exacerbations Before Enrollment

DECREASE IN	MEAN±SD	MEDIAN	RANGE	P VALUE
Number of hospital admissions	0.58±1.68	1.0	−11.70 to 5.62	<0.01^a
Number of ED visits	0.43±2.19	0	−8.47 to 7.43	<0.01^a
Number of exacerbations	1.01±3.00	2.0	−11.25 to 9.43	<0.01^a

Significant at the 1% level.

ED, emergency department; SD, standard deviation.

The only factor observed to be significantly associated with a positive decrease in number of exacerbations is the presence of other diseases (diabetes mellitus, hypertension, or congestive heart failure) at enrollment. The percentage of patients with a positive decrease in number of exacerbations was higher among patients with disease(s) other than COPD than among those with only COPD (94/117 [80.3%] versus 170/252 [67.5%]; p=0.011).

In the general CCHT sample (1,133 patients), among the 764 COPD patients with no exacerbations in the year prior to enrollment in the CCHT program, 170 (22.2%) died, and among the 369 patients included in the study, 95 (25.8%) died. There was no significant difference in death rate between the excluded and included groups (p=0.19).

Discussion

To our knowledge, this study is the first to evaluate specifically telemonitoring in patients with COPD and factors associated with reduction in healthcare utilization. Our study has a large enough sample size and enough power to show clear benefits associated with the program. With a mean enrollment time of 491 days, our study population had the longest follow-up of all previous studies.^12,13,15 In this study we showed an association between participation in this program and remarkably significant reduction in ED and UC COPD-related visits. About 71% of patients with more than one exacerbation per year had a reduction in acute care facility visits, and hence healthcare utilization.

Number of exacerbations prior to enrollment was the only factor that was associated with the significance of reduction in acute care facility visits. Those with two or more exacerbations per year had about 58% reduction in hospital admissions and 43% reduction in ED/UC visits. This was almost double the rates for those with one or more exacerbations (comparison between results presented in Tables 1 and 2).

Presence of other chronic health illnesses—namely, congestive heart failure, diabetes mellitus, and hypertension—was also associated with positive decrease in number of acute care facility visits. The percentage of patients with a positive decrease in number of exacerbations is higher among the patients with disease(s) other than COPD than among those with only COPD (94/117 [80.3%] versus 170/252 [67.5%]; p=0.011), indicating possibly positive effects of the CCHT on control of other chronic illnesses.

Our study has some limitations, and results should be interpreted with caution taking in consideration those limitations. First, it is a retrospective study that collected data using on ICD-9 codes and pooled data from the VA system. There is a potential of very few missed exacerbations for those seen outside the VA system. We expect this issue to have a very minimal effect on the study if any because most veterans seek care in the same hospital and are established and monitored closely in our system once enrolled in the CCHT program. In addition, the nurse coordinator has reported exacerbations whenever they happened and brought patients to the acute care facility when needed. We also expect that if a patient had an outside physician and was seen in another hospital for exacerbations, if he or she had visits before and after enrollment in the program, the effects of this would be close negligible or even have a trend to underestimate our results (toward the null effect) because those patients were monitored for a slightly longer period after enrollment and possibly had more visits reported after enrollment.

Second, our data were collected from patients enrolled in the program in a geographically restricted area (South-Central United States) with a majority white male population who have good access to healthcare and possibly higher socioeconomic status, making the results of our study restricted to that population. A more diverse study may be needed to generalize the results to external populations. We assume that in the presence of good access to healthcare and medications and the presence of a supervising physician that the same effects would be observed.

Third, the use of bursts of steroids and antibiotics was not assessed in our study because of difficulty accessing data outside the study facility and because on frequent occasions those medications might be prescribed for other reasons apart from COPD. The same reason applied for spirometric values to assess the magnitude of severity of COPD in those patients; in this case we accepted number of exacerbations as a surrogate, and, in fact, that might have been more accurate because there are many other factors, apart from forced expiratory volume in 1 s, that affect exacerbation frequency.⁶

Fourth, although our sample size was large enough to observe a significant impact of the CCHT program in reduction in health care utilization among COPD patients with one or more exacerbations, it was not large enough to determine if other factors are significantly associated with the reduction other than number of exacerbations before enrollment and presence of other diseases. We would have observed more factors to be noticed if the sample size had been larger.

Finally, the exact cost analysis was not assessed in our study because of factors such as the shared costs of CCHT for COPD with other chronic health conditions. Our study was not designed to assess cost; however, our program in general is a low-cost program and does not include home visits or home intervention arm. Previous reports indicated a significantly lower annual cost for CCHT compared to home-based primary care and nursing home care.¹⁶ We expect the program to be associated with significant savings as it is associated with reduced expensive ED visits and hospital admissions.

We believe that combining patient education about disease course, enrollment in a simple monitoring program like the CCHT that facilitates communications between patients and their healthcare professionals, and patient self-management would be a very efficient way to deal with chronic illnesses such as COPD and avert preventable hospital admissions or ED visits.

Conclusions

For patients with advanced COPD and frequent exacerbations, enrollment in a home telemonitoring program is associated with a reduction in acute care facility visits and should be considered in all patients with frequent exacerbations and in those with COPD as well as other chronic illnesses.

Footnotes

Acknowledgment

Dr. S. Alrajab contributed to the study design, data analysis, and manuscript composition and literature review; Dr. T.R. Smith contributed to study design, data collection, and institutional board review application; Dr. M. Owens contributed to manuscript review and study design; Dr. J.P. Areno contributed to study design and supervision; and Dr. G. Caldito contributed to statistical analysis and manuscript review.

Disclosure Statement

No potential conflicts of interest exist with any companies/organizations whose products may be discussed in this article.

References

Lopez

. Global burden of disease and risk factors. Oxford: Oxford University Press, 2006.

Foster

et al. Assessment of the economic burden of COPD in the U.S.: A review and synthesis of the literature. COPD, 2006; 3:211–218.

Welter

et al. Efficacy and tolerability of budesonide/formoterol added to tiotropium in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med, 2009; 180:741.

Bourbeau

et al. Reduction of hospital utilization in patients with chronic obstructive pulmonary disease: A disease-specific self-management intervention. Arch Intern Med, 2003; 163:585–591.

Bodenheimer

et al. Improving primary care for patients with chronic illness. JAMA, 2002; 288:1775–1779.

Cao

et al. Frequent hospital readmissions for acute exacerbation of COPD and their associated factors. Respirology, 2006; 11:188–195.

Joseph

. Care coordination and telehealth technology in promoting self-management among chronically ill patients. Telemed J E Health, 2006; 12:156–159.

Sicotte

et al. Effects of home telemonitoring to support improved care for chronic obstructive pulmonary diseases. Telemed J E Health, 2011; 17:95–103.

Paré

et al. Cost-minimisation analysis of a telehomecare program for patients with chronic obstructive pulmonary disease. Telemed J E Health, 2006; 12:114–121.

10.

Trappenburg

, Niesink

, de Weert-van Oene

, van der Zeijden

, van Snippenburg

, Peters

, Lammers

, Schrijvers

. Effects of telemonitoring in patients with chronic obstructive pulmonary disease. Telemed J E Health, 2008; 14:138–146.

11.

Dang

, Ma

, Nedd

, Aguilar

, Roos

. Differential resource utilization benefits with internetbased care coordination in elderly veterans with chronic diseases associated with high resource utilization. Telemed J E Health, 2006; 12:14–23.

12.

Bartoli

, Zanaboni

, Masella

, Ursini

. Systematic review of telemedicine services for patients affected by chronic obstructive pulmonary disease (COPD) Telemed J E Health, 2009; 9:877–883.

13.

Bolton

, Waters

, Peirce

, Elwyn

. Insufficient evidence of benefit: A systematic review of home telemonitoring for COPD. J Eval Clin Pract, 2010; 17:1216–1222.

14.

Smith

, Edwards

, Areno

, Walter

. Impact of telemedicine on healthcare utilization in patients with COPD. Chest Meeting Abstracts, 2009; 13692S-shttp://meeting.chestpubs.org/cgi/content/abstract/136/4/92S-a?maxtoshow=&hits=10&RESULTFORMAT=&author1=smith+&andorexactfulltext=and&searchid=1&FIRSTINDEX=0&sortspec=relevance&resourcetype=HWCIT. 2012 February 15.

15.

Barnett

et al. The cost-utility of a care coordination/home telehealth programme for veterans with diabetes. J Telemed Telecare, 2007; 13:318–361.

16.

Darkings

et al. Care coordination/home telehealth: The systematic implementation of health informatics, home telehealth, and disease management to support the care of veteran patients with chronic conditions. Telemed J E Health, 2008; 14:1118–1126.