The Role of Health Technologies in Multicomponent Primary Care Interventions: Systematic Review

Introduction

Primary care is often considered a cornerstone of health care systems. Health systems with strong primary health care produce better and more equitable health outcomes, are more efficient, and can achieve higher user satisfaction in comparison to health systems with only a weak primary care orientation [1,2]. Changing demographics, an increasingly aging population, and the increased burden of noncommunicable diseases have been identified as new challenges for health systems worldwide [3-5], and strengthening primary care has been proposed as one solution to address these challenges.

Many countries have implemented a wide array of innovations to enhance primary care, ranging from policy initiatives, such as capitated reimbursement, to ground level improvements, such as improving access to primary care practices and enhancing the roles of nurses to provide comprehensive primary care services [6-8]. As in other fields, such as finance, retail, and agriculture, an increasingly important domain for innovation involves the incorporation of technology. Technologies are having an impact on health service delivery and health system administration, and they promise to provide solutions for improving primary care [9,10].

There have been many studies emphasizing individual digital technologies for improving specific aspects of health care and primary care. Some of these include digital health assistants to help with administrative tasks, medical chatbots to engage patients more frequently, and the use of electronic health records and telemedicine, among others [9-11]. However, no studies have explored the role of technologies within multicomponent efforts to enhance primary care, that is, whether within initiatives comprised of several features aimed at enhancing primary care, there was a technology element being introduced, and if yes, what it was.

We aimed to systematically explore the role that health/digital technologies have played in multicomponent efforts designed to improve primary care by identifying (1) the types of technologies implemented, (2) the functional objective of the technology, (3) the relevant stakeholders, and (4) whether they have an impact on enhancing the defining features of primary care (ie, first contact, comprehensiveness, coordination, and continuity) [12], denoted here as the “4Cs.” We explored the overall outcomes of the multicomponent interventions in which technology is one component to attempt to discern the specific contribution of the technologies within these efforts. Textbox 1 provides useful definitions for concepts and terms that will be used throughout the article.

Methods

A systematic review was designed and performed following Cochrane guidelines for conducting systematic reviews [13]. The detailed methods for this review are described in an article that explored multicomponent interventions aimed at enhancing primary care, which identified 18 innovation strategies and provided a broader picture of the many innovation features used internationally to improve various aspects of primary care simultaneously [14]. A summary is provided below.

An electronic database search was performed in order to identify (1) multicomponent interventions or “innovation environments” aimed at enhancing primary care (with at least three innovation features); (2) factors influencing at least one of the primary care core functions (4Cs), and (3) studies reporting on any of the four basic types of outcomes of a successful health system (the so-called “quadruple aim” outcomes of population health, health care costs and utilization, patient satisfaction, and provider satisfaction) [15] and providing numerical values for at least five outcome measures. In a previous scan of the literature, we identified many specific interventions aimed at enhancing a particular aspect of primary care services (eg, the paper stamp checklist tool enhances asthma guideline knowledge and implementation by primary care physicians), and based on this, we determined that consideration of studies describing interventions with at least three distinct innovation features and measuring at least five outcome measures could ensure that the interventions were indeed “multicomponent.”

A search strategy was developed, and it focused on the following three main sets of terms: (1) primary care–related terms; (2) innovation/reform/enhancement-related terms; and (3) study design filters (Multimedia Appendix 1). The electronic database search was performed in Ovid/MEDLINE on May 30, 2019, and it was supplemented by manual searches through the references of the included studies and by a grey literature search (ie, search through materials and documents produced by organizations outside of the traditional commercial or academic publishing and distribution channels, such as government and industry documents) in OpenGrey [16], using “primary care” and “innovation,” on December 12, 2019. From the studies fulfilling these criteria, we selected those that had technology-based enhancements as part of the elements in their multicomponent interventions.

We defined health technologies, using definitions from two World Health Organization reports, as the “application of scientific knowledge for practical purposes, including its corresponding machinery and equipment, to solve health care–related problems and improve quality of life” [17] and encompassing digital health technologies (the overarching term to include eHealth and mHealth, eg, telemedicine, electronic health records [EHRs], and wearable sensors) and their corresponding medical and assistive devices [9].

Quality evaluation of the included studies was based on study design, using the National Institutes of Health–National Health, Lung and Blood Institute’s “Study Quality Assessment Tools” [18], a comprehensive suite of study evaluation tools, which has been used in a variety of systematic reviews [19-21]. Data extraction was performed using a predefined data extraction form for study characteristics and general information (author/year, setting/country, policy influence, study design and quality, and patient population involved), primary care intervention elements, and quadruple aim outcomes, including reported magnitudes for each outcome measure. A narrative descriptive approach was utilized to identify and report the types and specific details of the implemented technologies, the involved stakeholders, whether and which 4Cs were arguably supported, and the outcomes influenced by the corresponding technology.

Results

Search Results

After the electronic search, subjecting the articles to the inclusion/exclusion criteria and manual reference and grey literature searches resulted in 37 articles fulfilling the requirements for multicomponent interventions as described above. From these, 14 studies had technology-based enhancements and were included in the analysis (Figure 1).

‎View this figure

Study Characteristics

Articles were published between 2008 and 2017, and half of them were published since 2016. Most described studies performed in the United States (9/14, 64%). Additionally, four were from Europe (two from Germany and two from Spain) and one was from Argentina. Eight articles mentioned policies influencing the implementation of the innovation programs as broader country, regional, or organizational efforts to enhance primary care (Table 1) [22-35].

In terms of study designs and quality evaluation results, three publications reported controlled interventions (two of “fair” and one of “good” quality), six reported observational cohort or cross-sectional studies with controls (four of “fair” and two of “good” quality), one reported a case-control study of “fair” quality, and four reported pre-post studies without controls (one of “poor”, one of “fair,” and two of “good” quality). Populations studied or linked to the results included the general population enrolled in the programs (in six articles), chronically ill patients with one disease or complex chronic patients (in four articles), and special populations, including elderly and disadvantaged populations (in four articles).

The interventions in the articles included between four and 11 “innovation features” (see Multimedia Appendix 2 for definitions). The average number of features per intervention was seven (median seven), and the most common types, beside technology-based enhancements (present in all interventions), were innovations to improve access (in 11 articles), payment-based enhancements (in nine articles), and care/case management (in seven articles). In terms of the types of outcomes, the most commonly reported was health care costs and utilization (in 10 articles), followed by population health outcomes (in four articles), patient satisfaction (in three articles), and provider satisfaction (in one article). These are not mutually exclusive as one article reported on three outcomes and two reported on two outcomes each. The remaining 11 articles reported on one outcome each.

Technology-Based Results

Of the 37 articles, 14 (38%) describing multicomponent interventions to enhance primary care included technology-based enhancements as one of the innovation elements (Table 2) [22-35].

According to the descriptions of the articles, we were able to identify the following six broad categories for the types of implemented technologies (description below includes intended stakeholder and use):

1. 1. Enhancements leveraging electronic medical/health records [22,24-28,34]: it was the most common category (reported in seven studies from five interventions) and was aimed at providers. Their use is related to identifying specific groups of patients (eg, chronically ill) or specific needs of patients (eg, services needed), exchanging patient information, and developing quality measurements/control and risk stratification.
2. 2. Data monitoring technologies/online registries [23,26-29,34]: it was the second most common category (in six studies from four interventions) and was aimed at providers and practices. It was related to the management of utilization data to allocate resources and improve care, help with population-based management, and check on the impact of programs.
3. 3. Web-based online portals and messaging platforms [27,28,32,34,35]: it was included in five studies (reporting on three interventions) and was aimed at patients to access their medical records, obtain additional health promotion information, promote self-management, and facilitate access and communication with providers.
4. 4. Medication-specific eHealth/information technology tools [30,31]: it was included in two studies and was aimed at providers to support pharmacotherapy and medication prescription.
5. 5. Telephone-based enhancements [33]: it was described in one article and was aimed at providers to communicate with patients and caregivers, and provide follow-up to reduce patients’ nonessential clinic visits.
6. 6. Electronic consultations between providers [33]: it was described in one study and was aimed at enhancing geriatric referrals.

Based on the description of the technological enhancements included in the studies, we were able to link them to the 4Cs in the following way:

1. 1. First contact: Three programs aimed to apply technology to impact this feature through telephone-facilitated access to primary care colleagues, facilitated appointment scheduling through web portals, and updated digital health systems for easier patient access.
2. 2. Comprehensiveness: Six interventions sought to increase the ability to manage a wider range of problems with technology, including providing additional health promotion information through patients’ web portals and enhancing capacity for providers to better use electronic medical records, improve medication prescription, and provide improved preventive and chronic care.
3. 3. Coordination: Three programs used technology to improve care coordination by improving EHR-enabled information exchange and by allowing electronic consultations to facilitate care among primary care providers and specialists.
4. 4. Continuity: Nine interventions sought to enhance the longitudinal relationship between patients and providers by enhancing the identification and follow-up of patients for individualized care, allowing more comprehensive identification and monitoring of service needs, and improving communication between patients and providers via online messaging or telephone contact.

Outcomes

Since these technology-based innovation elements are part of wider innovation environments, which include additional enhancement features, it was not possible to attribute outcomes specifically to the identified technologies. However, we still present the outcomes of the full innovation environments in an effort to elucidate the potential role of these technologies in the outcomes. The numerical magnitudes for each outcome are presented in Multimedia Appendix 3 (along with details of the full intervention). Table 2 and the paragraphs below present a descriptive summary and general direction of the effects for these outcomes.

Overall, the studies presented mixed results (ie, nonsignificant changes or significant benefits and deteriorations simultaneously for a specific outcome) for all types of outcomes, except for provider satisfaction, which was reported only in one study. The most consistent improvements per type of outcome were as follows: (1) health care costs and utilization, increased cost saving and decreased costs for some parameters (eg, Medicare expenditure decreased by US $11 per beneficiary per month [25] and drug prescriptions decreased by €44 per patient [31]) and increased primary care visits compared to specialists; (2) population health, improved blood pressure control, improved glycated hemoglobin, decreased amputations and end-stage renal disease, and decreased smoking status; (3) patient satisfaction, increased satisfaction with timely appointments and self-management support and increased satisfaction with the ability to see the usual doctor; (4) provider satisfaction, improved perception toward place of work’s quality and services provided, and its consideration as a good place to work.

The most consistent mixed results by the type of outcome were as follows: (1) health care costs and utilization, nonsignificant changes or simultaneous improvements and deteriorations depending on the study for hospital admissions, readmissions, and emergency department visits; (2) population health outcomes, nonsignificant changes for cholesterol and lipid levels, myocardial infarction, and stroke; (3) patient satisfaction, no differences for communication with providers and for primary care provider performance.

The most consistent deteriorations were found for some health care costs and utilization outcomes, such as increased number of prescriptions, increased costs for general practitioner (eg, intervention €27 more expensive than control per patient) and specialist (intervention €22 more expensive than control per patient) consultations [31], and increased costs for prescription coverage.

Discussion

Principal Findings

Only 38% of our identified multicomponent interventions that aimed at enhancing primary care included technology-based enhancements, highlighting the fact that technology has not played a major role in comprehensive efforts aimed at enhancing primary care. This is not surprising, as it has been widely acknowledged that innovation in health care has always been difficult [36], especially if it has involved digital or technological efforts [37-39].

Most of the included articles reported on health care costs and utilization outcomes, signaling that technology-based efforts are either aimed at decreasing costs and utilization or at least not increasing costs without contributing to other aspects of system success. In fact, the only considerable unintended consequences were increased costs for general practitioner and specialist visits, and increased costs of prescription coverage (in some studies), suggesting that introducing technologies in health care can lead to increased costs, as it has been consistently reported in the literature [40-42].

The most common technology identified within these efforts was EHR, which is also not surprising given the widespread advocacy for this technology [43,44], and it was aimed mainly at providers or practices to facilitate information exchange among them and improve monitoring efforts. The only identified technology aimed at patients was the deployment of online patient portals, where they can see their records, message their providers, and access additional health information mostly for health promotion, which is in line with the idea that patients are ever more active participants in their own health care [44,45].

When analyzing the interventions in terms of their impact on the 4Cs, the technologies implemented were mostly aimed at improving continuity by increasing the identification and follow-up of patients (with labels in EHRs and telephone communication), enhancing monitoring efforts for identifying care and service needs (also mostly through EHRs and online registries), and ensuring more constant communication between providers and patients via online messaging. This reflects the growing importance of continuity of care, which in the past has had weak evidence linked to its benefits, but was recently highlighted as important, especially with regard to its link to decreased mortality risk [46,47]. Technologies have been promoted to improve comprehensiveness by providing additional health promotion information for patients, improving the ability of providers to prescribe medications, reinforcing the ability of primary care providers to cover a broader number of issues themselves, and avoiding overreferring [47].

In terms of outcomes, the literature provides limited but useful information. For example, increased patient satisfaction with timeliness of care, scheduling, and better self-management support could be in part explained by the use of online patient portals. Such portals allow patients to schedule appointments, see their own medical records, and access additional prevention information. Increased primary care visits, relative to specialist visits, appear to result from innovations that enhance monitoring of services needed and follow-up of patients (identified through EHRs and/or by telephone follow-up). The introduction of medication-specific digital/information technology tools could be associated with differing impacts. While studies reported a decrease in the costs of drug prescriptions and medication therapies, they also reported an increase in the mean number of drugs prescribed, and it was also associated with more costly consultations (around €25 extra per consultation [31]).

In order for digital technologies to play a more prominent role in primary care enhancement efforts, there is first a need for a responsible policy to support their development and introduction [48]. For example, some of the primary care enhancement environments have included an explicit policy encouraging the introduction of technology or information technology initiatives as part of their efforts [25,31,49]. To make this happen successfully, the technology must be seen as a tool that provides needed functions in a way that is effective, humane, and sustainable. Here, the context in which the technology will be implemented must be considered. It is essential to engage relevant stakeholders to deeply understand their environment and capabilities so that the introduced technology will be truly useful, improve (or at least not disrupt) existing workflow, and have tangible value [37,50]. In order to establish value, there is a need for technologies to be linked to evidence-based positive outcomes, taking into account their potential to improve health outcomes, costs, and patient and provider satisfaction.

There are some limitations for this study. The nature of the search and the specific requirements for including studies (ie, those describing multicomponent interventions aimed at enhancing primary care, which provided numerical magnitudes for reporting quadruple aim outcomes) may have made us overlook other important technological innovations aimed at improving primary care that had qualitative assessments only or did not measure quadruple aim outcomes. Similarly, since this review only focused on published and grey literature, it did not account for quality improvement interventions implemented at, for example, private primary care or accountable care organizations, which may have included technological innovations but no published results. Therefore, although technology may have a more prominent role in primary care–enhancing initiatives overall, the results from published and grey literature do not indicate so. Additionally, the fact that technological enhancements were one of many components within a primary care enhancement effort, our study eligibility criteria did not allow us to establish the actual and specific impact of the technologies on outcomes. However, it did help to situate these technologies within multicomponent innovation strategies and to gain preliminary insights into how technological enhancements may support other nontechnologically based innovation features and their impacts on the four primary care functions.

Taking all of this into account, future research should try to pinpoint the specific impact of technology-based innovation features within wider efforts aimed at enhancing primary care. This would mean including specific measures that could link outcomes to the use of such technology and quantify this impact. This exercise would also help to identify which of the 4Cs of primary care is being impacted by this technology, which would help understand the mechanisms of how these innovations are improving care. An additionally interesting research direction would be to explore how technological innovations are being leveraged at primary care private practices and/or accountable care organizations, to understand the actual role of technology in quality improvement initiatives for which there is no publicly available data. Such research could provide a more balanced view of the actual usage of technological innovations in primary care at the ground level.

Conclusions

Although technology and digital health have been proposed and encouraged as possible solutions to improve primary care, they have not played a major role in formally evaluated multicomponent interventions aimed at enhancing primary care, as reflected in the published and grey literature. Other types of nontechnologically based innovations, such as those aimed at improving access, restructuring payments for providers, and providing team-based care, have been much more widely implemented, reflecting that digital health technologies have not yet reached maturity or wider acceptance as a means for improving primary care. Leveraging technologies already in use, such as EHRs, and internet-based technologies, such as online patient portals, seems to provide promising avenues to improve continuity and comprehensiveness in primary care, which may eventually lead to better health outcomes and improved patient satisfaction. A stronger push is needed if technologies are meant to support wider efforts aimed at enhancing primary care and for them to play a more substantial role within these efforts. High-level policy and financial support must be designed to focus on the needs of a diversity of stakeholders and to encourage evidence-based research based on a coherent set of methods and measures. In this way, we can hope to fulfill the promise of technologies and digital health to enhance health care through strong sustainable primary care.