TITLE: Personal Electronic Health Records: A Review of Clinical … · Personal Electronic Health Records 2 educational material, 4and other features designed to promote patient self-management.-7

Disclaimer: The Rapid Response Service is an information service for those involved in planning and providing health care in

Canada. Rapid responses are based on a limited literature search and are not comprehensive, systematic review s. The intent is to provide a list of sources of the best evidence on the topic that CADTH could identify using all reasonable efforts w ithin the time allow ed. Rapid responses should be considered along w ith other types of information and health care considerations. The

information included in this response is not intended to replace professional medical advice, nor should it be construed as a recommendation for or against the use of a particular health technology. Readers are also cautioned that a lack of good quality evidence does not necessarily mean a lack of effectiveness particularly in the case of new and emerging health technologies, for w hich little information can be found, but w hich may in future prove to be effective. While CADTH has taken care in the preparation

of the report to ensure that its contents are accurate, complete and up to date, CADTH does not make any guarantee to that ef fect. CADTH is not liable for any loss or damages resulting from use of the information in the report. Copyright: This report contains CADTH copyright material and may contain material in w hich a third party ow ns copyright. This

report may be used for the purposes of research or private study only. It may not be copied, posted on a w eb site, redistributed by email or stored on an electronic system w ithout the prior w ritten permission of CADTH or applicable copyright ow ner.

Links: This report may contain links to other information available on the w ebsites of third parties on the Internet. CADTH does not have control over the content of such sites. Use of third party sites is governed by the owners’ own terms and conditions .

TITLE: Personal Electronic Health Records: A Review of Clinical Effectiveness, Cost-Effectiveness, and Guidelines

DATE: 04 March 2016

CONTEXT AND POLICY ISSUES

Chronic disease, such as diabetes, chronic obstructive pulmonary disease and circulatory diseases, is one of the most common and burdensome health problems worldwide. Canadians are living longer and the aging population is more likely to experience more chronic conditions. On the other hand, there is a trend towards chronic diseases occurring earlier in life; the prevalence rates of chronic diseases are increasing faster among Canadians aged 35 to 64 years than those older than 65 years.1 For instance, the prevalence of diagnosed diabetes among Canadians increased by 70% from 1998-1999 to 2008-2009.2 The greatest relative increase in prevalence was seen in the 35 to 39 and 40 to 44 year age groups, where the proportion doubled.2 In Ontario, 62% of women and 55% of men reported having at least one chronic disease, and 29% of Ontario adults aged 25 and older reported having 2 or more chronic diseases in 2005.3 In Canada, management of chronic disease accounts for 67% of all direct healthcare costs and costs the economy $190 billion annually for treatment and lost productivity.1 Given the increasing prevalence of common chronic diseases and the associated costs, effective and efficient chronic disease management is essential for the Canadian healthcare system.

Patients with a chronic disease often have multiple concurrent chronic conditions and complications that require regular visits with a number of different specialists in addition to their primary care physician, or may have a chance to be treated in the emergency department or other acute care settings for disease progression or severe adverse events. Timely and accurate health information exchange between healthcare providers is essential for chronic disease management. A variety of electronic tools have been developed to enhance information sharing between various healthcare providers, and sometimes patients themselves are involved.4 Electronic personal health records (PHR) or patient portals are patient-accessible systems which are operated via the internet. Patients may be able to manage and share their health information with healthcare providers in a private, secure and confidential environment. In addition, PHR and patient portals provide secure interactions between patient and physician, remote disease monitoring, prescription requests and/or renewals, access to higher quality

Personal Electronic Health Records 2

educational material, and other features designed to promote patient self-management.4-7 In general, patient portals differ from PHR in terms of ownership. The data in a patient portal are owned and managed by physicians and hospitals along with the electronic health record (EHR). The advantage of a portal over a PHR is that the data are updated whenever there are updates on the EHR, while the data in a PHR are only updated when the patient updates them.4,8 It has been suggested that when patients are actively involved in disease management (e.g., have access to their medical records and secure communications with healthcare providers), their health outcomes can be improved.9 However, previous studies have yielded inconsistent results for the clinical benefits of PHR or patient portals.7 The purposes of this review is to provide evidence on the comparative clinical effectiveness and cost-effectiveness of personal electronic health records, and to summarize the guidelines that are relevant to the use of personal electronic health records in chronic disease management. RESEARCH QUESTIONS

1. What is the clinical effectiveness of personal electronic health records for patients

managing an active chronic health condition?

2. What is the cost-effectiveness of personal electronic health records for patients managing an active chronic health condition?

3. What are the evidence-based guidelines regarding the use of personal electronic health records by patients managing an active chronic health condition?

KEY FINDINGS

There was insufficient evidence to directly demonstrate the effect of personal electronic health records on important health outcomes. Improved laboratory results such as lower blood pressure, decreased HbA1c levels and decreased cholesterol levels were reported in the intervention group; however, a statistically significant difference was not always detected, or a clinically significant difference was not observed between treatment groups. On the other hand, the use of personal electronic health records is related to better patient-provider communication, improved patient satisfaction and improved self-disease management. More economic evidence is needed to demonstrate the cost-effectiveness of personal electronic health records. METHODS Literature Search Methods

A limited literature search, with main concepts appearing in title, abstract or major subject heading, was conducted on key resources including PubMed, Ovid Medline, The Cochrane Library, University of York Centre for Reviews and Dissemination (CRD) databases, Canadian and major international health technology agencies, as well as a focused Internet search. No filters were applied to limit the retrieval by study type. Where possible, retrieval was limited to the human population. The search was also limited to English language documents published between January 1, 2011 and February 4, 2016. Rapid Response reports are organized so that the evidence for each research question is presented separately.


Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 1.

Table 1: Selection Criteria Population Patients managing an active chronic condition

Intervention Q1 to Q3: Personal electronic health records

Comparator Q1 and Q2: in-person appointments; paper-based or phone-based

communication with healthcare providers Q3: No comparator necessary

Outcomes Q1: Clinical effectiveness and harms Q2: Cost-effectiveness Q3: Guidelines regarding the use of personal electronic health records

Study Designs Health technology assessments (HTAs), Systematic reviews (SRs), Meta-analyses (MAs), Randomized controlled trials (RCTs), Non-randomized controlled trials, Economic evaluations, Evidence-based guidelines.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 1, they were duplicate publications, or were published prior to 2011. HTAs, MAs, and SRs were excluded if there was incomplete reporting of methods or if they were superseded by a more recent or more rigorous review. Clinical trials were excluded if they were included in a selected systematic review. Studies were also excluded if the investigated intervention was not patient-accessible or lacked patient-provider interaction. Critical Appraisal of Individual Studies

The included systematic reviews were critically appraised using the AMSTAR checklist,10 randomized and non-randomized controlled studies were critically appraised using the Downs and Black checklist,11 and economic studies were assessed using the 35-item Drummond’s checklist.12,13 Summary scores were not calculated for the included studies; rather, a review of the strengths and limitations of each included study were described narratively. SUMMARY OF EVIDENCE Details of study characteristics, critical appraisal, and study findings are located in Appendices 2, 3, and 4, respectively.


Quantity of Research Available

A total of 633 citations were identified in the literature search. Following screening of titles and abstracts, 585 citations were excluded and 48 potentially relevant reports from the electronic search were retrieved for full-text review. Three potentially relevant publications were retrieved from the grey literature search. Of these potentially relevant articles, 40 publications were excluded for various reasons, while 11 publications met the inclusion criteria and were included in this report. Appendix 1 describes the PRISMA flowchart of the study selection. Summary of Study Characteristics

Study Design Five SRs,14-18 three RCTs,19-21 one non-RCT9 and two economic evaluations22,23 were identified. All of the SRs restricted their literature search to English language publications only. The majority of the included individual studies in the SRs had a non-experimental study design. An SR conducted by Kruse and colleagues

15 was an update of a previous SR published in 2011.

Compared with the earlier review, the Kruse et al. review adopted wider selection criteria (observational studies and existing systematic reviews were also examined while the previous review included experimental studies only). The purpose of the Price et al. review16 was to identify which conditions were potentially sensitive to the PHR as an intervention and would benefit from PHR-enabled management. All three included RCTs had an open-label study design, due to the nature of the intervention.19-21 The non-RCT was a controlled before-and-after study.9 Two economic evaluations estimating the cost-effectiveness of patient portals in patients with chronic diseases were identified.22,23 The first report was a cost-effectiveness analysis conducted in Finnish public primary care in 2012, using data from a non-randomized trial9 by the same authors.22 The perspective of this analysis was not specified. Direct costs related to the patient portal use during a 6-month period were estimated. Nonparametric bootstrapping was used to simulate 1000 incremental cost-effectiveness ratios (ICERs). The second economic evaluation was a cost-utility analysis from the perspective of the Ontario Ministry of Health and Long-term Care. Clinical data from an RCT conducted in Ontario were used as inputs into a computerized simulation model (the Ontario Diabetes Economic Model) to estimate diabetes-related complications (e.g., myocardial infarction, amputation and renal failure), life-expectancy, quality-adjusted life-year (QALY) and the costs of complications in patients with type 2 diabetes. Direct medical costs related to the diabetic complications in the year in which an event occurred as well as the costs in subsequent years associated with the ongoing management of the complication were used in the model. There were no evidence-based guidelines identified with guidance of use of patient portal or PHR in patients with chronic diseases identified. Country of Origin The SRs were conducted in Canada16 and the United States.14,15,17,18 The RCTs were conducted in Canada21 and the United States.19,20 The non-RCT was conducted in Finland.9 The two economic evaluations were conducted in Canada23 and Finland.22


Patient Population In the Bush review,14 studies on children with asthma were included. The number of patients in the studies for this review ranged from 7 to 84,015. The other SRs involved adult patients with various chronic conditions, such as diabetes, hypertension, hyperlipidemia and HIV infection.15-18 The RCTs were conducted in children with asthma19 or adult patients with diabetes,21 hypertension,20,21 hypercholesterolemia,21 coronary artery disease21 or other cardiovascular diseases.21 The non-RCT enrolled patients with at least two chronic conditions.9 The Finnish economic evaluation22 included patients with chronic diseases such as hypertension, diabetes and hypercholesterolemia, while the Canadian economic evaluation23 examined the cost-effectiveness of the intervention specifically in diabetic patients. Interventions and Comparators The clinical effectiveness of patient portals was assessed in four SRs,14,15,17,18 while PHR was assessed in another SR.

16 There was a clear distinction between PHR and patient portal based

on the ownership of these two technologies. In addition, the functions of patient portals varied in different reports. In the SRs, the patient portals investigated in all of the included individual studies enabled patients to access their own health records such as treatment plan and lab results, receive educational information, and communicate with the health care providers via secure messaging. In some studies, patients might be allowed to upload data to the portals, to receive health maintenance reminders, to refill prescriptions, to schedule medical appointments, or to pay bills. The interventions investigated in the RCTs and non-RCT were all patient portals linked to EHR.9,19-21 EHR was maintained by the health care provider or hospital. The studies did not distinguish patient portals from PHR in terms of data ownership. The comparators in the clinical trials were all usual care. In the non-RCT, patient outcomes before and after the study were also compared in each group.9 The cost studies included patients using electronic messaging services from patient portals,22 or patients with computerized decision support system (CDSS) linked to electronic medical records (EMRs).23 The comparator in both studies was usual care where there was no patient portal or CDSS. Outcomes The SRs evaluated the utilization of patient portals or PHR, patients’ satisfaction, barriers and facilitators to the use of patient portals or PHR, and their impact on health outcomes which were measured by disease-specific lab results in the study populations.14-18 In the included RCTs, the effectiveness of patient portals on the improvement in clinical outcomes such as change in blood pressure and glucose levels, missed school/work days due to the disease was evaluated, as well as the feasibility and/or acceptability of patient portals in the study populations.19-21 In the non-RCT, the effect of access to patient portal on patient activation (measures patients’ knowledge of their disease, skills to self-manage their disease, and self-confidence in their abilities to manage their disease) was evaluated using a validated questionnaire, the short form of Patient Activation Measure (PAM13).9


The Finnish economic evaluation calculated the ICER and cost-effectiveness acceptability curves (CEACs),22 based on patient activation and health status. The Canadian economic evaluation calculated the incremental cost per QALY, the extra monetary resources needed for the intervention to gain one extra life year (not quality-adjusted). Harms were not reported in any of the included studies. Summary of Critical Appraisal

In all the included SRs, objectives were clearly described. Literature searches were restricted to English language publications. Duplicate study selection and data extraction procedures were indicated in three SRs.14-16 The trial characteristics and the patient characteristics of the included individual studies were not adequately reported. The quality of the included studies was low, due to the study design of these studies, or insufficient outcome reporting. The majority of the studies were retrospective, which raised the potential for confounding factors impacting the study conclusions. There was no assessment of publication bias, and there was considerable heterogeneity in the interventions (different electronic personal health record systems) and demographics among the included studies. Authors of all these SRs reported potential conflicts of interest, or provided funding sources. All of the included RCTs had an open-label design due to the nature of the interventions. The baseline patient characteristics were comparable between the intervention groups and the control groups. In one trial, data collectors, medical record reviewers/ telephone interviewers were blinded to group allocation; data analysis and manuscript preparation were kept blinded until the end of analysis.21 Loss to follow-up was reported in all RCTs. The intention-to-treat principle was not followed in two RCTs.19,20 Power calculations or methods of sample size determination were not reported in one trial.20 In the non-RCT, patients with at least two chronic conditions were enrolled and assigned to the intervention group or the control group based on their date of birth.9 The patient characteristics were similar between groups at baseline. A power calculation was not provided. Patient baseline activation scores were a covariate in the analysis of covariance; however, it is unclear whether other potential confounders had been controlled for in the analysis. Results should be interpreted with caution due to the small sample size and short study duration (the 6-month duration may not be long enough to detect meaningful changes in clinical effectiveness outcomes). This study has limited generalizability, because the study population and the investigated intervention may not be similar to a Canadian population that would use patient portals. One of the economic evaluations was conducted from the perspective of the Ontario Ministry of Health, using Canadian cost data and the clinical inputs from a Canadian clinical trial. The base case assumed a 1-year treatment effect, 5% discount rate and 40-year time horizon (which may be sufficient for chronic diseases). Its results are likely to be generalized to a Canadian patient population, although the data in this study may be somewhat outdated (published in 2012 and using 2010 Canadian dollars).23 In this study, sources for costs were provided for the intervention and alternatives. Besides the base case analysis, sensitivity analyses were performed according to the changes in treatment duration and discount rates. The results of the analyses were sensitive to change in treatment duration. Insufficient details were presented in a Finnish economic evaluation.22 The main assumptions seemed reasonable but there was a lack of data regarding time horizon, perspectives, discount rates, and a sensitivity analysis was not performed.


Summary of Findings

What is the clinical effectiveness of personal electronic health records for patients managing an active chronic health condition? Systematic reviews

Patient portals were the most commonly investigated intervention in the included SRs.14,15,17,18 PHR was examined in one SR.16 The effects of patient portals on clinical outcomes were mixed: there was a trend that patient portals were associated with lower HbA1c levels, lower low-density lipoprotein (LDL), lower total cholesterol or lower blood pressure. However, the statistically significant differences between users and non-users were not always achieved, or the differences were not always considered clinically important. Inconsistent results on the number of emergency department (ED) visits/ primary care visits/ specialist visits/ hospitalization were reported in the SRs. Patient satisfaction was improved after the use of patient portals in two SRs,14,15 but mixed results were observed in the third.

18 Patients indicated that patient portals were easy to use, accurate and timely, and

improved patient-provider communication, but still for some patients, they couldn’t replace phone calls. Barriers to the adoption of patient portals were similar across the SRs: lack of internet use, lack of health literacy, difficulty in understanding medical terminology, need for assistance in explanation of lab test results, lower socioeconomic status, older age, and gender (however the reported gender differences were inconsistent across reports) were commonly reported. In terms of facilitators, younger age, higher trust in the internet, higher computer literacy, family member or healthcare provider recommendation, and non-Hispanic /non-black race positively influenced the utilization of patient portals. In one SR, the associations between PHR and the health outcomes of a range of chronic diseases were evaluated. The findings suggested that patients in the tethered PHR group (PHR was connected to another clinical information system such as a regional EMR/EHR) were more likely to report benefit, compared with those in the standalone PHR group (PHR was not connected to another clinical information system). Seven conditions were identified as potentially PHR-sensitive with patients potentially benefiting from this intervention: asthma, diabetes, fertility, glaucoma, HIV infection, hyperlipidemia and hypertension. Randomized controlled trials

Patient portals linked to electronic health records were investigated in all RCTs.

In one RCT enrolling children with asthma,19 patient portal-users experienced fewer ED visits, hospitalizations, specialist visits, and primary care visits. Compared to children in the comparison group, use of patient portals was associated with fewer lost school days for the children (difference was not statistically significant) and fewer lost work days for the parents (difference was statistically significant). The frequency of asthma flare-ups was significantly lower in the intervention group. The difference in quality of life between groups was not statistically significant.

Findings from another RCT20 enrolling patients with uncontrolled hypertension suggested that 6-month use of a patient portal was related to lower risks of hypertension, diabetes, and hyperlipidemia; however the differences between users and non-users were not statistically


significant, except for the change in weight lost from baseline. There was no statistically significant difference in the quality of life between users and non-users.

A third RCT evaluated the effect of patient portal use in patients with previous vascular events.21 The authors indicated that even though favorable and significant results from the use of patient portal were observed in the complex processes required to reduce vascular risks, its effect on clinical outcomes, including quality of life, was not significant.

Non-randomized controlled trial

Results from a Finnish study9 enrolling 137 chronically ill patients suggested that after six months of patient portal use, patients in the intervention group showed improvements in knowledge about their diseases and skills of self-management, and gained more confidence in managing their condition.

What is the cost-effectiveness of personal electronic health records for patients managing an active chronic health condition?

Economic evaluations The literature search identified two economic evaluations on the cost-effectiveness of personal electronic health records in patients with chronic disease. In the Canadian study,23 the base case analysis in diabetic patients found that one year use of a shared computerized decision supporting system resulted in a relative risk reduction of 14% in the occurrence of amputation compared with the control group, and led to an ICER of C$156,970 per life-year gained and C$160,845 per QALY gained. In a Finnish study,22 the cost associated with the portal use increased by an average of €48; at a willingness to pay €700 per clinically significant change in patient activation, there was a greater than 50% chance that the intervention was cost-effective; at a willingness to pay €2100 per clinically significant change in patient activation, there was a 70% chance that the intervention was cost-effectiveness. However, the results may not be generalized to a Canadian population. What are the evidence-based guidelines regarding the use of personal electronic health records by patients managing an active chronic health condition? There were no evidence-based guidelines identified in this report. Limitations One limitation in the included studies (trials and systematic reviews) is the heterogeneity of the interventions. Patient portals varied considerably in the formats, which included providing access to test results or medical notes, or secure messaging. The interventions also had various features in the included studies, for example, using patient portals to access handouts or self-management information, to record and track subjective experience data or objective data related to the condition over time, to drive evidence-based reminders and alerts to the user to support self-management, to make day-to-day decisions about care management such as medication dosing, or to communicate with the care team and support group.16 Data from the included systematic reviews were inadequately reported, such as the patient characteristics in the individual studies by many trials which precluded classification of risks of


bias. Furthermore, the majority of the individual studies were not able to provide compelling evidence to demonstrate the clinical benefit from patient portal or PHR, due to the low study quality resulting from the study design (surveys, focus groups, interviews and observational studies). The majority of the reported results were related to patient satisfaction, improvements in self-management, access to care, access to information and sense of control. These are typically self-reported by patients or health care providers. There is a lack of objective data, or use of a validated or standardized reporting tool to support such findings. Utilization of the interventions is a difficult concept to quantify. Different methods were employed in the included studies to measure utilization, such as the frequency of accessing the patient portal or PHR, number of secure messages sent through the portal or PHR, number of glucose uploads and self-reported usage.17 Therefore, it is challenging to compare the results across different studies. In addition, given the nature of most of the evidence in this review (studies with non-experimental design), there is an issue that people who are active patient portal users may also have higher-than-expected clinical need or be more active in other aspects of their self-care; therefore this may be a reason for the differences of the observed trends. The cost-effectiveness studies had limited data sample sizes and there was a high degree of heterogeneity of costs between the included studies due to differences in costs of technologies and in healthcare systems across countries. CONCLUSIONS AND IMPLICATIONS FOR DECISION OR POLICY MAKING

Although the clinical effectiveness of personal electronic health records in patients with chronic diseases were evaluated in a number of systematic reviews, randomized and non-randomized controlled studies, there was insufficient evidence to directly demonstrate its effect on important health outcomes such as reduction in rate of cardiovascular disease, mortality or health-related quality of life. There was a trend of improved laboratory results (such as lower blood pressure, decreased HbA1c levels and decreased cholesterol levels) related to the use of personal electronic health records; however, a statistically significant difference was not always detected, or a clinically significant difference was not observed between treatment groups. On the other hand, the use of personal electronic health records is related to better patient-provider communication, improved patient satisfaction and improved self-disease management. Barriers to the use of this technology include, but are not limited to, age, lack of internet access, poor health literacy and sociodemographic disparities. Even though the existing clinical evidence shows improved short-term risk factors for chronic diseases, more economic evidence is needed to demonstrate the cost-effectiveness of personal electronic health records and promote patient self-management in the study population. PREPARED BY:

Canadian Agency for Drugs and Technologies in Health Tel: 1-866-898-8439 www.cadth.ca

http://www.cadth.ca/


REFERENCES

1. Elmslie K. Against the growing burden of disease [Internet]. Ottawa: Public Health Agency

of Canada; 2012. [cited 2016 Feb 25]. Available from: http://www.ccgh-csih.ca/assets/Elmslie.pdf

2. Diabetes in Canada: facts and figures from a public health perspective [Internet]. Ottawa: Public Health Agency of Canada; 2011 Dec 15. [cited 2016 Feb 25]. Available from: http://www.phac-aspc.gc.ca/cd-mc/publications/diabetes-diabete/facts-figures-faits-chiffres-2011/highlights-saillants-eng.php

3. Optimizing Chronic Disease Management in the community (outpatient) setting (OCDM): an evidentiary framework. Ontario Health Technology Assessment series [Internet]. 2013 Sep [cited 2016 Feb 23];13(3):1-78. Available from: http://www.hqontario.ca/Portals/0/Documents/eds/ohtas/evidentiary-framework-ocdm-130906-en.pdf

4. Electronic tools for health information exchange: an evidence-based analysis. Ontario Health Technology Assessment series [Internet]. 2013 Sep [cited 2016 Feb 23];13(11):1-76. Available from: http://www.hqontario.ca/Portals/0/Documents/eds/ohtas/full-report-etools-health-information-exchange-130906-en.pdf

5. Tenforde M, Jain A, Hickner J. The value of personal health records for chronic disease management: what do we know? Fam Med [Internet]. 2011 May [cited 2016 Feb 11];43(5):351-4. Available from: http://www.stfm.org/fmhub/fm2011/May/Mark351.pdf

6. Emont S. Measuring the impact of patient portals: what the literature tells us [Internet]. Oakland (CA): California HealthCare Foundation; 2011 May. [cited 2016 Feb 23]. Available from: http://www.chcf.org/~/media/MEDIA%20LIBRARY%20Files/PDF/PDF%20M/PDF%20MeasuringImpactPatientPortals.pdf

7. Tenforde M, Nowacki A, Jain A, Hickner J. The association between personal health record use and diabetes quality measures. J Gen Intern Med [Internet]. 2012 Apr [cited 2016 Feb 11];27(4):420-4. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3304034/

8. Kruse CS, Argueta DA, Lopez L, Nair A. Patient and provider attitudes toward the use of patient portals for the management of chronic disease: a systematic review. J Med Internet Res [Internet]. 2015 [cited 2016 Feb 11];17(2):e40. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4376181/

9. Riippa I, Linna M, Ronkko I. The effect of a patient portal with electronic messaging on patient activation among chronically ill patients: controlled before-and-after study. J Med Internet Res [Internet]. 2014 [cited 2016 Feb 11];16(11):e257. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4260064/

http://www.ccgh-csih.ca/assets/Elmslie.pdfhttp://www.ccgh-csih.ca/assets/Elmslie.pdfhttp://www.phac-aspc.gc.ca/cd-mc/publications/diabetes-diabete/facts-figures-faits-chiffres-2011/highlights-saillants-eng.phphttp://www.phac-aspc.gc.ca/cd-mc/publications/diabetes-diabete/facts-figures-faits-chiffres-2011/highlights-saillants-eng.phphttp://www.hqontario.ca/Portals/0/Documents/eds/ohtas/evidentiary-framework-ocdm-130906-en.pdfhttp://www.hqontario.ca/Portals/0/Documents/eds/ohtas/evidentiary-framework-ocdm-130906-en.pdfhttp://www.hqontario.ca/Portals/0/Documents/eds/ohtas/full-report-etools-health-information-exchange-130906-en.pdfhttp://www.hqontario.ca/Portals/0/Documents/eds/ohtas/full-report-etools-health-information-exchange-130906-en.pdfhttp://www.stfm.org/fmhub/fm2011/May/Mark351.pdfhttp://www.chcf.org/~/media/MEDIA%20LIBRARY%20Files/PDF/PDF%20M/PDF%20MeasuringImpactPatientPortals.pdfhttp://www.chcf.org/~/media/MEDIA%20LIBRARY%20Files/PDF/PDF%20M/PDF%20MeasuringImpactPatientPortals.pdfhttp://www.ncbi.nlm.nih.gov/pmc/articles/PMC3304034/http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4376181/http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4260064/


10. Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, et al. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol [Internet]. 2007 [cited 2016 Mar 4];7:10. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1810543/

11. Downs SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health [Internet]. 1998 Jun [cited 2016 Mar 4];52(6):377-84. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1756728/

12. Drummond MF, Richardson WS, O'Brien BJ, Levine M, Heyland D. Users' guides to the medical literature. XIII. How to use an article on economic analysis of clinical practice. A. Are the results of the study valid? Evidence-Based Medicine Working Group. JAMA. 1997 May 21;277(19):1552-7.

13. Drummond MF, Jefferson TO. Guidelines for authors and peer reviewers of economic submissions to the BMJ. The BMJ Economic Evaluation Working Party. BMJ [Internet]. 1996 Aug 3 [cited 2016 Mar 4];313(7052):275-83. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2351717

14. Bush RA, Connelly CD, Fuller M, Perez A. Implementation of the Integrated Electronic Patient Portal in the Pediatric Population: A Systematic Review. Telemed J E Health. 2015 Aug 10.

15. Kruse CS, Bolton K, Freriks G. The effect of patient portals on quality outcomes and its implications to meaningful use: a systematic review. J Med Internet Res [Internet]. 2015 [cited 2016 Feb 11];17(2):e44. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4342639/

16. Price M, Bellwood P, Kitson N, Davies I, Weber J, Lau F. Conditions potentially sensitive to a personal health record (PHR) intervention, a systematic review. BMC Med Inf Decis Mak [Internet]. 2015 [cited 2016 Feb 11];15:32. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4411701/

17. Amante DJ, Hogan TP, Pagoto SL, English TM. A systematic review of electronic portal usage among patients with diabetes. Diabetes Technol Ther. 2014 Nov;16(11):784-93.

18. Goldzweig CL, Orshansky G, Paige NM, Towfigh AA, Haggstrom DA, Miake-Lye I, et al. Electronic patient portals: evidence on health outcomes, satisfaction, efficiency, and attitudes: a systematic review. Ann Intern Med. 2013 Nov 19;159(10):677-87.

19. Fiks AG, Mayne SL, Karavite DJ, Suh A, O'Hara R, Localio AR, et al. Parent-reported outcomes of a shared decision-making portal in asthma: a practice-based RCT. Pediatrics. 2015 Apr;135(4):e965-e973.

20. Green BB, Anderson ML, Cook AJ, Catz S, Fishman PA, McClure JB, et al. e-Care for heart wellness: a feasibility trial to decrease blood pressure and cardiovascular risk. Am J Prev Med [Internet]. 2014 Apr [cited 2016 Feb 11];46(4):368-77. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3978093/

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1810543/http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1756728/http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2351717http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4342639/http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4411701/http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3978093/


21. Holbrook A, Pullenayegum E, Thabane L, Troyan S, Foster G, Keshavjee K, et al. Shared electronic vascular risk decision support in primary care: Computerization of Medical Practices for the Enhancement of Therapeutic Effectiveness (COMPETE III) randomized trial. Arch Intern Med. 2011 Oct 24;171(19):1736-44.

22. Riippa I, Linna M, Ronkko I. A Patient Portal With Electronic Messaging: Controlled Before-and-After Study. J Med Internet Res [Internet]. 2015 [cited 2016 Feb 11];17(11):e250. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4642411

23. O'Reilly D, Holbrook A, Blackhouse G, Troyan S, Goeree R. Cost-effectiveness of a shared computerized decision support system for diabetes linked to electronic medical records. J Am Med Inform Assoc [Internet]. 2012 May [cited 2016 Feb 11];19(3):341-5. Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3341787/

24. Piette JD, Striplin D, Marinec N, Chen J, Gregory LA, Sumerlin DL, et al. Improving Post-Hospitalization Transition Outcomes through Accessible Health Information Technology and Caregiver Support: Protocol for a Randomized Controlled Trial. J Clin Trials [Internet]. 2015 Oct [cited 2016 Feb 11];5(5). Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4711915/

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4642411http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3341787/http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4711915/


APPENDIX 1: Selection of Included Studies

585 citations excluded

48 potentially relevant articles retrieved for scrutiny (full text, if

available)

3 potentially relevant reports retrieved from other sources (grey

literature, hand search)

51 potentially relevant reports

40 reports excluded: -irrelevant population (4) -irrelevant intervention (18) -irrelevant outcomes (4) -other (review articles, editorials) (10) -already included in at least one of the selected systematic reviews (4)

11 reports included in review

633 citations identified from electronic literature search and

screened


APPENDIX 2: Characteristics of Included Publications

Table A1: Characteristics of Included Health Technology Assessments and Systematic Reviews

First Author,

Publication Year,

Country

Types and numbers

of primary studies included

Patient

Characteristics

Intervention(s) Comparator(s) Clinical

Outcomes

Bush 2016,

United States

14

11 studies

investigating the utilization and impact of patient portal.

Number of patients ranged from 7 to 84,015.

Literature search: from 1992 to August 2014;

English only. Type of studies was not specified.

Patients 0-18 years of

age; with chronic condition.

Patient portal: a secure web

site, integrated with the EHR though which patients can complete forms, communicate

electronically with healthcare providers, access personal health information, schedule

appointments, request prescription refills and pay bills.

Not specified. The

included studies were survey, interview or

observational studies.

Portal utilization,

patient satisfaction, barriers for the

portal use, and clinical outcome.

Kruse 2015,

United States

15

27 studies (RCTs,

non-RCTs, observational studies, SR) summarized the

results on effect of patient portals on chronic-condition

outcomes. Literature search: from

January 2011 to August 2014. This review was an update

of a previous SR published in 2011.

Patients with chronic

conditions.

Patient portal (defined as “a

secure online website that gives patients convenient 24-hour access to personal health

information from anywhere with an Internet connection. Patients had access to their health

information”).

No portal use in the

included experimental studies.

No comparators in the descriptive studies or surveys.

Chronic condition

outcomes, patient satisfaction, barriers,

Price 2015, Canada

16

23 studies (including 7 RCTs) were included

to identify the

Patients with various chronic conditions, in

outpatient

PHR (defined as “electronic health records controlled,

shared or maintained by

No PHR Care quality, access,

productivity.



First Author, Publication

Year,

Country

Types and numbers of primary studies

included

Patient Characteristics

Intervention(s) Comparator(s) Clinical Outcomes

conditions which were sensitive to PHR intervention.

Literature search: from 2008 to August 2014.

English only.

environment. 10 conditions were found from the literature

search.

patients to support patient centered care”. -17 studies on tethered PHR

(connected to EHR) -6 studies on standalone PHR (not connected to EHR)

Amante 2014, United States

17

16 studies were included to evaluate the patient portal use

in diabetic patients. Number of patients ranged from 20 to

15,427. Literature search: from

February 2005 to January 2014. English only and only articles

from studies within the United States were included.

Patients with diabetes Patient portal (online personal health record that is tethered with a healthcare provider’s

electronic health record system).

No patient portal use Utilization, barriers and facilitators of

patient portal use

Goldzweig 2013, United

States18

46 studies were included to evaluate

patient portal use in patients with various conditions. Number of

patients ranged from 9 to 88,642.

Literature search: from 1990 to January 2013. A separate search

Patients with chronic condition such as

diabetes, hypertension, heart failure, infertility or

depression. The disease was not specified in some

studies.

Patient portal (tethered to existing health care institutions

as opposed to those standalone).

No patient portal use Health outcomes, patient

satisfaction, adherence, efficiency,

utilization, attitudes, and patient

characteristics.




Year,

Country

Types and numbers of primary studies

included

Patient Characteristics

Intervention(s) Comparator(s) Clinical Outcomes

exclusive for PHR and patient portal was conducted from

November 2011 to January 2013. English publications only.

EHR=electronic health record; PHR=personal health record; RCT=randomized controlled trial; SR=systematic review.


Table A2: Characteristics of Included Clinical Studies First Author, Publication

Year, Country, Study Name

Study Design, length of

follow-up

Patient Characteristics Intervention(s) Comparator(s) Main clinical outcomes

RCTs

Fiks 2015, United States

19

RCT, 6-month follow-up.

Convenient sample of 60

families (30 families in each group).

Randomization was stratified by

practice and by asthma severity level.

Children 6-12 years of age, with persistent

asthma, from both urban and suburban practices in 2 states.

MyAsthma: EHR-linked SDM portal.

Features included identification of patents’ concerns

and treatment goals, tracking of symptoms/ drug side

effects/ progress toward goals; educational content;

and access to the child’s asthma care plan.

Usual care - Feasibility of portal (% of participants in the

intervention group who completed a survey during the study);

- Acceptability of asthma care (11 Likert-scaled questions, and questions

re. patient satisfaction); - Clinical outcomes (number of ED visits/

hospitalizations/ specialist visits/ primary care visits, number of

prescriptions for asthma medications, number of lost school/ work days,

asthma-related QOL)

Green 2014, United States

20

RCT, 6-month follow-up.

101 patients were randomized

(intervention group: 51; control group:

50).

Patients aged 35-69 years with BMI > 26, systolic BP > 140 mmHg or diastolic

BP > 90 mmHg, and Framingham 10-year CVD risk score between 10%

and 25%, who were registered patient website users, and were EHR data

identified

Web-based dietitian-led team care intervention (follow-

up occurred via secure messaging to report BP, weight

and diet, and received ongoing feedback).

Usual care (received a copy of their lab results including Framingham 10-

year CVD risk via the patient website and by mail. They received no

other active treatment).

Change in BP, weight loss ≥ 4 kg at 6-month, % of patients with BP

control, intervention utilization, satisfaction.

Holbrook 2011, Canada

21

RCT, 1 year follow-up.

1102 patients

Patients ≥ 55 years, with previous vascular events (myocardial infarction,

angina or coronary artery

CDSS intervention (a web-based individualized

vascular tracking

Usual care Change in PCS (calculated as the sum of the frequency-weighted

process score for each of


Table A2: Characteristics of Included Clinical Studies First Author, Publication

Year, Country, Study Name

Study Design, length of

follow-up

Patient Characteristics Intervention(s) Comparator(s) Main clinical outcomes

were randomized

(intervention group: 545; control group:

557).

disease, stroke or transient ischemic attack or

peripheral vascular disease), diabetes, hypertension or

hypercholesterolemia, in community-based primary care setting.

and advice-decision support system

integrated with EMR).

the 8 main risk factors*; change in CCS based on

the same 8 risk factors; QOL.

Non-RCTs

Riippa 2014,

Finland9

Non-randomized

controlled before-and-after study, 6-month

follow-up. Patient allocation was

based on their date of birth.

N=137 (intervention: 80; control 57)

Chronically ill patients in

public primary care in Finland, ≥ 18 years of age, had ≥ 2 treatable health

conditions

Patient portal

(including patients’ medical records, care plan and

secure messaging with a care team).

Patients in this group received immediate access to

the patient portal, and received their care plan through

the portal.

Usual care.

Patients in this group received delayed portal

access after 6 months, but received a printed copy of their care plan.

Patient activation

(measured by PAM13, which assesses a patient’s knowledge of

their diseases, skills to self-manage their disease, and self-

confidence in managing their condition)

BMI=body mass index; BP=blood pressure; CCS=clinical composite score; CDSS=computerized decision support systems; CVD=cardiovascular disease; ED=emergency department; EHR=electronic health record; EMR=electronic medical record; PCS=process composite score; PAM13=short form of Patient Activation Measure; PCCP=patient centered care plan; QOL=quality of life; RCT=randomized controlled trial. *the 8 main risk factors were: blood pressure, LDL levels, weight, aspirin or equivalent therapy, smoking, exercise, diet, and psychosocial index; the total possible score for PCS was 27.


Table A3: Characteristics of Included Economic Evaluations First Author, Publication

Year, Country

Study Design, Time horizon

Patient Characteristics Intervention/ Comparators

Assumptions

Riippa 2015, Finland

22*

Cost-effectiveness analysis, using data from a clinical trial

with 6-month follow-up.

Time horizon was not reported.

137 patients with at least 2 chronic conditions (80 in the

intervention group, 57 in the control group).

Intervention: patient portal including patient medical records, care

plan, and secure messaging with a care team.

Control: usual care without the use of

patient portal.

A secure message to the patient using the patient portal was considered to require the same amount of a health care professional’s

time as sending an email or a letter. The cost of providing access to the patient portal

for this study was €6/year/patient - based on a 5-year depreciation plan of the portal deployment cost, on the expected average

number of users during this period, and on the yearly maintenance cost.

The number of patient portal users in 2014 2 years after portal adoption was 3527, and the number was expected to grow by 1000 users

per year in 2015 and 2016.

O’Reilly 2012, Canada

23

Cost-utility analysis. Time horizon: 40

years. Discount rate: 5%

for both costs and effects.

Perspective: the Ontario Ministry of Health and Long-

term Care.

511 patients with type 2 diabetes from an RCT conducted in Canada

(253 in the intervention group, 258 in the control group).

Intervention: a shared CDSS linked to EMR (web-based,

continuously updated, secure, patient-specific diabetes tracker).

Control: usual care without the use of CDSS

linked to EMR.

In the base case, it was assumed that there were no continuing benefits of the program in terms of its impact on risk factors beyond the

one-year intervention period; the treatment effect was assumed to last for only 1 year.

There were different assumptions regarding the duration of the program and treatment effect such as 5 and 10 years.

CDSS=computerized decision support system; EMR=electronic medical records; ICER=incremental cost-effectiveness ratio; QALY=quality-adjusted life year; RCT=randomized controlled trial. *More clinical data in this economic evaluation are presented in Riippa 2014

9, Table A2.

Personal electronic health records 20

APPENDIX 3: Critical Appraisal of Included Publications

Table A4: Strengths and Limitations of Health Technology Assessments and Systematic Reviews using

AMSTAR24

Strengths Limitations

Bush 201614

Objectives were explicit

Literature search performed on multiple databases

Study selection and data extraction were

conducted by independent reviews

List of included studies was provided

Quality of the included studies was evaluated

Conflict of interest was reported

No review protocol published prior to conduct of review

Not clear if grey literature search was conducted; search strategy was not provided in details

Search restricted to English language publications

Most data were collected from surveys, focused

interviews or case series; no experimental studies were included; no longitudinal data

Sufficient details of the included studies were not provided, such as patient’s disease,

Quality of the evidence was low due to the study designs of the included individual studies

Publication bias was not examined.

Kruse 201515

Objectives were explicit, research protocol was developed prior to the study

Comprehensive literature search


Intervention, comparator and outcomes were explicit

Risk of bias of the included studies was assessed

Study selection, data extraction and quality assessment of the included studies were performed by 2 independent reviewers

Conflict of interest was reported.

Unclear if the literature search was restricted to English language publications articles only

Publication bias was not assessed.

Price 201516



Intervention, comparator and outcomes were

explicit

Study selection and data extraction of the included studies were performed by 2 independent reviewers



Unclear if a research protocol has been developed prior to the study

Search restricted to English language

publications

Types of study were not specified in search strategy

Study results were reported as “”improved disease control” but no further details for disease-specific outcomes such as HbA1C level, blood pressure and mortality

Quality assessment of the included studies was not conducted

Publication bias was not examined.

Amante 201417



Intervention, comparator and outcomes were

Unclear if a research protocol has been developed prior to the study

Search restricted to English language


Table A4: Strengths and Limitations of Health Technology Assessments and Systematic Reviews using AMSTAR

24


explicit

Quality assessment of the included studies was performed


publications and articles from studies within the United States

Types of study were not specified in search strategy

Data extraction was performed by one reviewer; it is unclear if study selection and quality

assessment were performed by two reviewers independently

Study results were reported as “”improved

disease control” but no further details for disease-specific outcomes such as HbA1C level, blood pressure and mortality

Quality assessment of the included studies was not reported

Publication bias was not examined

Goldzweig 201318



Intervention, comparator and outcomes were explicit; some core functionalities of patient portal, including access to test results, secure messaging, and disease management

information, were presented in almost all systems to ensure the comparability of the interventions across the studies

Key questions, search strategy and inclusion/ exclusion criteria were developed with input from a technical expert panel

Quality assessment of the included studies was performed


Search restricted to English language publications

it is unclear if study selection and data extraction were performed by two reviewers independently

Patient characteristics were not described in details; patient’s condition was not specified in

some studies

Publication bias was not assessed


Table A5: Strengths and Limitations of Clinical Trials using the Downs and Black Checklist11


RCTs

Fiks 201519

• Objectives and inclusion/ exclusion criteria were stated. • Patient characteristics, interventions, and

outcomes were described. • Lost-to-follow up was reported • Conflict of interest was declared

• Baseline participant’s characteristics were comparable between groups

• Method used to generate the randomizat ion sequence was unclear • Sample size calculation was not described;

differences were described but the statistical significance was not tested • Intent-to-treat analysis was not performed

• Convenient sample based on clinician recommendation – the study was not designed to be representative of all children with asthma in the

care network. • Funding source was not declared.

Green 201420

• Objectives and inclusion/ exclusion criteria were stated.

• Patient characteristics, interventions, and outcomes were described. • Choice of sample size was justified.

• Lost to follow up was reported. • Baseline participant’s characteristics were comparable between groups, except for gender

(23% more women in the patient portal group). • Funding sources were disclosed.

• Method used to generate the randomization sequence was unclear.

• Intent-to-treat analysis was not performed • QOL was assessed, but an MCID was not reported to determine the clinical relevance.

• Generalizability limited; uncertain as to whether study patients were representative of all patients.

Holbrook 201121

• Objectives and inclusion/ exclusion criteria were stated.

• Patient characteristics, interventions, and outcomes were described. • Randomization sequence was generated by an

allocation-concealed online program. • Data collectors, medical record reviewers/ telephone interviewers were blinded to group

allocation; data analysis and manuscript preparation were kept blinded until the end of analysis.

• Sample size calculations were described • Lost to follow up were reported • Intent-to-treat analysis was used with multiple

imputation for missing data. • Conflict of interest was declared.

• Although the intervention was labelled as “web-based”, 28.7% of all study participants reported that

they never used the Internet. • The composite scores used in this study have not been validated, and an MCID was not indicated.

• Generalizability limited; uncertain as to whether study patients were representative of all patients

Non-RCTs

Riippa 20149

• Objectives and selection criteria were stated. • Patient characteristics, interventions, and

outcomes were described. • Lost to follow up was reported. • A validated tool was used to measure patient

activation. • No statistically significant differences in patient’s baseline characteristics.

• Conflict of interest was declared.

• Sample size calculation was not described • Generalizability limited; uncertain as to whether

study patients were representative of all patients.


ICER=incremental cost-effectiveness ratio; MCID=minimal clinically important difference; QOL=quality of life

Table A6: Strengths and Limitations of Economic Evaluations using the Drummond

Checklist12,13 Strengths Limitations

Riippa 201522 • Purpose of the study was clearly described. • Research question was clearly described. • Comparator was appropriately defined.

• Provided detailed information on clinical inputs such as effectiveness. • Resource use and costs were described.

• Propensity-score matching was used to adjust for the baseline differences between groups. • Conflict of interest was declared.

• The 6-month follow-up period was short to examine the benefit of the intervention. • Discount rate was not reported.

• Perspective was not described. • Values of ICER were not reported. • Sensitivity analysis was not performed.

• The study was conducted using euro cost information from Finland which may limit the generalizability to Canada.

O’Reilly 201123 • Research question was clearly described. • Clinical inputs such as effectiveness were provided in details.

• Resource use and costs were described and justified • Perspective, discount rate and time horizon were

described • Sensitivity analyses were conducted. • The study was conducted in Canadian healthcare

setting; therefore the results are likely to be generalizable to Canadian population. • Conflict of interest was declared.

• This study was conducted 5 years ago and the costs were in 2010 Canadian dollars.


APPENDIX 4: Main Study Findings and Author’s Conclusions

Table A7: Summary of Findings of Included Systematic Reviews

Main Study Findings Author’s Conclusions

Bush 201614

11 studies were included, all were observational studies:

utilization rates in 4 studies: registration was generally low, 4% in patients of 0-16 years of age and 65% in a survey of parents of children 0-5 years of age;

patient satisfaction in 6 studies: the patient portal was easy to use (92%), accurate and timely, but it still couldn’t replace phone calls to healthcare providers.

barriers in 6 studies: difficulty in graphing and data

interpretation, needing more explanation of test results and medical terminology, and confidentiality.

Patient characteristics in 4 studies: users differed from

nonusers in ethnics, income levels, education and with/without commercial insurance

Clinical outcomes in 1 study: portal use was statistically significant predictor of glycosylated hemoglobin level, but

not of low-density lipoprotein and total cholesterol levels.

Results from observational studies

suggested that patient portal was easy to use, and may be related to improvement in glycosylated

hemoglobin level; but there were still barriers for its widely use.

Further studies are needed to confirm its impact on health

outcomes.

Kruse 201515

27 studies were included (26 experimental or non-experimental studies and 1 review).

In 10/27 studies reported improvements in medication

adherence, disease awareness, self-management of disease, ↓of office visits, ↑in preventative medicine, ↑in patient satisfaction and increased patient-to-provider

communication among patient portal users. Weak results were reported on medical outcomes.

Barrier identified for patient portal adoption: lack of internet use, lack of technical support, lack of health literacy and numeracy skills (portal users could not accurately interpret

lab results).

More health care organizations today offer features of a patient portal than in the review published

in 2011

Several clinically relevant benefits were identified in portal users.

The number of office visits were observed

No sufficient evidence to demonstrate its efficacy.

Price 201516

23 studies were included (7 of them were RCTs) 76% (13 out of 17 studies with used tethered PHRs) reported

benefit; 50% (3 out of 6 studies with standalone PHRs) reported benefit.

Clinical benefit from PHR were reported for these 7 conditions: -Asthma: improved ability to self manage, more activated patient;

-Diabetes: improved patient satisfaction/ disease control/ communication with healthcare provider/ access to patient information;

-Fertility: improved continuity/ access to health knowledge/ patient-provider communication; -Glaucoma: improved medication management;

-HIV: more activated patient, improved disease control/ access to own information/ access to health knowledge/ patient-provider communication/ ability to self-manage;

7 conditions were identified as potentially PHR-sensitive; benefits were seen in care quality, access

and productivity

Common characteristics of the PHR-sensitive conditions: chronic,

the self-management behaviors could be suitably tracked in a PHR and were tightly linked to the

feedback of monitoring/self-monitoring of indicators, e.g. self-monitored blood pressure in

hypertension, or glucose levels in diabetes




-Hyperlipidemia: improved ability to self-manage/ disease control; -Hypertension: improved ability to self manage, more activated

patient; PHR had no benefit for cancer, idiopathic thrombocytopenic

purpura and multiple sclerosis.

Amante 201417

16 experimental or non-experimental studies were included. Factors related to higher portal enrollment:

Higher education, younger age, higher income and non-Hispanic/non-black race, have access to computer/ internet, private health insurance, higher self-efficacy, more diabetes

knowledge, higher health literacy and better-controlled diabetes. Factors related to higher portal utilization:

Higher education, younger age, higher income and non-Hispanic/non-black race. Inconsistent results were found for the factors of: higher health literacy, diabetes-related factors.

Barriers to portal use: lack of patient capacity/ desire/ awareness of portal functions, and negative patient experience

Facilitators to portal use: providers and family member recommendation.

Several demographic characteristics, having better-controlled diabetes, and providers who engaged in and encouraged

portal use were associated with increased portal enrollment and utilization

Major barriers to portal enrollment include inadequate access and trouble using technology

Major facilitators to portal utilization: accessible technology, trust, physician and family

encouragement, knowledge of the technology.

Goldzweig 201318

46 studies were included: 14 RCTs, 21 observational studies and 11 descriptive studies.

Health outcomes, satisfaction and adherence (12 RCTs, 6 observational studies): -Mixed results were reported for HbA1c levels, blood pressure

or low-density lipoprotein cholesterol levels (no statistically significant difference between groups although a trend of favoring the portal group was observed, or statistically

significant improvement observed in the patient portal group but the between group difference may not be clinically important); -Mixed results were reported in evaluation on patient satisfaction

between portal users and non-users; -General adherence to medical advice was improved in patient portal group in 1 observational study, but was lower in patient

portal group in another observational study. Efficacy/utilization (5 RCTs, 2 observational studies):

-Inconsistent results were reported in the rates of emergency department visits, primary care or specialty visits, and hospitalization;

Patient characteristics (9 observational studies, 1 descriptive study):

“Evidence that patient portals improve health outcomes, cost, or utilization is insufficient. Patient attitudes are generally positive, but

more widespread use may require efforts to overcome racial, ethnic, and literacy barriers." (pp.677)




-Age, gender, ethnics, socioeconomic status, access to internet, level of education and level of health literacy were found to be significantly associated with the patient portal use.

Attitudes, barriers and facilitators (2 RCTs, 6 observational studies, 11 descriptive studies):

-Younger patients were more likely to accept portal use and had more trust in the internet/computer literacy; -administrative and human factors in the interface were cited as

barriers to portal use. PHR=personal electronic health records; RCT=randomized controlled trial;


Table A8: Summary of Findings of Included Clinical Trials


Year, Country

Main Findings and Authors’ Conclusion

Randomized controlled trials

Fiks 2015,

United States19

Feasibility:

Frequent users (completed the portal survey ≥ 5 times during the 6-month period): 57%;

Satisfaction:

No significant differences in satisfaction between 2 groups (data not provided) During the 6-month period, asthma-related health care use: (Intervention group vs.

control group):

ED visits: 3 vs. 9

Hospitalizations: 0 vs. 2

Specialist visits: 11 vs. 21

Primary care visits: 29 vs. 41

Prescriptions for inhaled steroids per child: 1.1 vs. 0.7

Prescriptions for oral steroids per child: 0.4 vs. 1.0

Missed school days: -1.1 vs. -0.4, p=0.2

Missed work days: -1.1 vs. +0.7, p=0.001

Frequency of asthma flares: -3.3 vs. -1.3, p=0.02

QOL: Nighttime symptoms: 15.7 vs. 16.3, p=0.9 Daytime symptoms: 12.0 vs. 8.0, p=0.3

Functional limitations: 9.3 vs. 5.0, p=0.4

Authors’ Conclusion: After 6 months “use of an EHR-linked asthma portal was feasible and acceptable to families and improved clinically meaningful outcomes ”. pp.1

Green 2014,

United States20

At 6-month follow up (intervention vs. control):

Change in weight lost from baseline: -3.7 kg vs. -0.5 kg, p< 0.01;

Change in systolic BP from baseline: -13.9 mmHg vs. -11.4 mmHg, p=0.40;

Change in diastolic BP from baseline: -8.5 mmHg vs. -6.6 mmHg, p=0.32;

Change in CVD risk score from baseline: -4.1 vs. -2.8, p=0.1;

Change in HbA1c from baseline: 0 vs. 0.1, p=0.11;

Change in total cholesterol from baseline: -5.7 vs. -1.8, p=0.52;

Change in LDL from baseline: -7.0 vs. -5.2, p=0.71;

Change in HDL from baseline: 1.5 vs. -0.3, p=0.24;

Satisfaction with hypertension care (0-10 scale): 1.3 vs. -0.5, p


Table A8: Summary of Findings of Included Clinical Trials


Year, Country


vs. 20, p=0.02);

QOL as measured by the EQ-5D score: no significantly changed (data not shown);

Number of vascular events during the 1-y follow-up: 27 (5%) vs. 30 (5.4%), p=0.75;

Authors’ Conclusion: “Despite favourable ratings from patients in terms of continuity of care and self-efficacy

in handling their personal vascular risks and important improvements in the complex processes required to reduce vascular risk, clinical variables were not consistently improved.” pp. 1743

Non-randomized studies

Riippa 20149 At 6-month follow up:

Change in patient activation scores from baseline (intervention vs. control):

+1.05 vs. -1.58, difference 2.77 (95% CI -1.24 to 6.79); the difference was smaller than an MCID of 4-5 points;

Effect of time since last diagnosis on change in final patient activation score (overall population):

o diagnosed with a severe condition during the intervention: +5.4, SD8.4; o diagnosed 1-2 years ago: +2.3, SD15.7; o no severe diagnoses: +1.6, SD13.1;

o severe diagnosis made > 2 years before the intervention: -7.1, SD12.3; o diagnosed < 1 year before the intervention: -3.0, SD11.5.

Effect of baseline patient activation level on change in final patient activation

score (overall population): o low levels (1-2): +8.5, SD12.3; o level 3: +0.7, SD11.7;

o level 4: -6.1, SD11.3.

Authors’ Conclusion: “Time since last severe diagnosis and patient activation at baseline may affect changes

in patient activation, suggesting that these should be considered in evaluation of activating chronic care interventions and in the specification of possible target groups for these interventions”. pp.2

CI=confidence interval; ED=emergency department: EHR=electronic health record; HDL=high-density lipoprotein; LDL=low-density lipoprotein; MCID=minimal clinically important difference; QOL=quality of life; SD=standard deviation.


Table A9: Summary of Findings of Included Economic Evaluations


Year, Country


Riippa 201523

From adjusted model:

Costs increased by an average of €48; CEAC results showed at a willingness to pay €700 per clinically significant change in patient activation, there was a > 50% chance that the intervention was cost -

effectiveness; CEAC results showed at a willingness to pay €2100 per clinically significant change in patient activation, there was a 70% chance that the intervention was cost-

effectiveness;

Author’s conclusion: “Offering the possibility to substitute health care visits with less costly contracts using

self-management tools did not seem to compromise the health status or treatment of chronic care patients.” (p. 3)

O’Reilly 201123

Base case analysis (all costs were 2010 Canadian dollars)

Total cost of implementing the intervention was $483,699;

1-year intervention resulted in a relative risk reduction of 14% in the occurrence of

amputation in the intervention group compared to the control group;

The intervention resulted in an additional 0.0117 QALY;

The ICER was $156,970/life-year gained and $160,845/QALY gained.

Sensitivity analyses:

If the patients in the model were treated for 5 years, there was an increase in the number of incremental QALYs from 0.0117 in the base case to 0.0421. Thus the

increase in QOL and life years gained would be translated into an ICER of $186,728.

If the treatment period was extended to 10 years, the incremental cost -

effectiveness ratio was $173,654.

Author’s conclusion: “The web-based prototype decision support system slightly improved short -term risk

factors. The model predicted moderate improvements in long-term health outcomes. This disease management program will need to develop considerable efficiencies in terms of costs and processes or improved effectiveness to be considered a cost -

effective intervention for treating patients with type 2 diabetes .” (pp. 341)

CEAC=cost-effectiveness acceptability curve; ICER=incremental cost-effectiveness ratio; QALY=quality-adjusted life-year.

Context and policy issuesResearch questionSkey FindingsMethodsLiterature Search MethodsSelection Criteria and MethodsExclusion CriteriaCritical Appraisal of Individual Studies

Summary of EVIDENCEQuantity of Research AvailableSummary of Study CharacteristicsSummary of Critical AppraisalSummary of FindingsLimitations

Conclusions and implications for decision or policy makingReferences