Electronic Medical Record Adoption in Dutch Hospitals Barriers and Facilitators

2015

H.C. Vroege

University of Amsterdam June 2015

Electronic Medical Record

Adoption in Dutch Hospitals

Barriers and Facilitators

Barriers and Facilitators

Senior Implementation Consultant Furore

SRP Tutor

Stephanie Medlock

Klinische Informatiekunde (KIK) AMC - UvA

Location of Scientific Research Project Furore

Bos en Lommerplein 280 1055 RW Amsterdam

Practice teaching period November 2014 – June 2015

This Master thesis describes the Scientific Research Project (SRP) which I worked on during the last year of my study that I worked at Furore and is the final project of the Master Medical Informatics. Medical Informatics is a two year academic course which is taught at the Academic Medical Center (AMC), which hosts the medical faculty of the University of Amsterdam (UvA).

I would like to express my gratitude to everyone who has helped me with my SRP.

Special thanks go out to my mentor Richard Hopmans and my tutor Stephanie Medlock for all the time they invested in guiding me. I appreciate their constructive feedback, and their kindness. Without you I would not have been able to complete this project.

I also want to thank everyone I interviewed for their time and cooperation, all colleagues at Furore for their support and advice, and my fellow students for their support and the fun dinners.

Finally I want to give special thanks to my parents and sisters for always believing in me, motivating me, and supporting me; and to my friends for assisting me and distracting me from my SRP from time to time. Thank you for all encouraging words and endless support!

Summary ... 1

Samenvatting ... 2

Chapter 1 Introduction ... 3

Research objectives ... 3

Chapter organization ... 4

Chapter 2 Adoption of the Electronic Medical Record ... 5

Electronic Medical Record ... 5

EMR Adoption ... 5

Adoption models ... 6

Technology Acceptance Models ... 6

HIMSS Electronic Medical Record Adoption Model... 7

Gartner’s Generation Model ... 9

Chapter 3 Barriers and Facilitators to EMR Adoption A Literature Review ... 11

Introduction... 11

Methods ... 11

Information extraction ... 12

Taxonomy ... 13

Results ... 13

Study characteristics and quality. ... 14

Barriers and solutions ... 15

Facilitators to EMR adoption ... 21

Benefits of EMR adoption ... 22

Discussion ... 23

Chapter 4 EMR Adoption in Dutch Hospitals ... 24

Introduction... 24

Methods ... 24

Interview questions ... 24

Interviews and coding of results ... 24

Results ... 25

Interview questions ... 25

Interview results ... 25

Section 1 – Implementation strategy ... 26

Section 2 - Barriers and facilitators by domain in academic hospitals ... 27

Section 3 – Implementation process and EMR benefits ... 31

Comparison with literature ... 32

Discussion ... 33

Chapter 5 Discussion ... 35

References ... 38

List of Abbreviations ...40

H.C. Vroege 1 | P a g e

Summary

Hospitals throughout the world are looking for ways to keep improving the quality of care while containing or reducing costs. One of the ways in which hospitals in the Netherlands try to achieve this, is by improving their healthcare IT capabilities by implementing an

integrated new generation Electronic Medical Record (EMR). An EMR is optimally utilized if the entire organization uses the software. The decision of a health care professional to make full use of the EMR is called EMR adoption. The aim of this research is to find the barriers, solutions to barriers, and facilitators of EMR adoption. Furthermore we aim to determine how adoption models are being used in defining the implementation strategy and in what way hospitals and patients benefit from a higher EMR adoption rate. To support this aim, this thesis contains a literature review and a qualitative study consisting of semi-structured interviews.

The literature review led to the inclusion of twenty articles from PubMed, Web of Science, LISTA, Google Scholar, and HIMSS Analytics. We extracted barriers, solutions to barriers, facilitators, and benefits to EMR adoption. We extracted thirty barriers, seven solutions, and nine facilitators, and categorized them in a taxonomy.

With the interviews we wanted to capture three concepts: EMR adoption, domains of the EMR, and the implementation process and benefits. The interviewer recorded, transcribed, and coded each interview. We interviewed eighteen people involved in an EMR

implementation in a Dutch hospital, and extracted thirty-five barriers, three solutions, and thirty-seven facilitators to EMR adoption.

The barriers to EMR adoption in both the literature and in Dutch hospitals were mainly in the technical, time, and social categories. The solutions to barriers were mainly in the technical, psychological, legal, and change process categories. The facilitators were mainly in the technical and social categories. Two barriers and two facilitators were found in the interview results that were not mentioned in the literature. These barriers and facilitators are related to more advanced aspects of the EMR. These aspects are a clinical data warehouse,

medication administration, clinical decision support, and health information exchange. From the interviews we concluded that IT management rarely used adoption models in

determining their implementation strategy.

The benefits reported by hospitals with a high level of EMR adoption were an increase in the efficiency of care processes, an increase in quality, and safety of care. In addition to these benefits, our interview subjects anticipate facilitating patient empowerment through the use of patient portals. The findings of this study can be applied to help achieve these benefits by increasing EMR adoption. From this work we can conclude that numerous barriers exist that could prevent a higher level of EMR adoption. This study includes possible solutions to overcome these barriers and provides an overview of facilitators that can be used to increase EMR adoption.

Keywords

H.C. Vroege 2 | P a g e

Samenvatting

Ziekenhuizen in de hele wereld zoeken naar manieren om de kwaliteit van zorg te blijven verbeteren en tegelijkertijd de kosten van zorg gelijk te houden of te verlagen. Een van de manieren waarop Nederlandse ziekenhuizen dit proberen te bereiken is door de IT

voorzieningen te verbeteren door de implementatie van een geïntegreerd nieuwe generatie Elektronisch Patiënten Dossier (EPD) te implementeren. Een EPD wordt maximaal benut als de volledige organisatie gebruik maakt van de software. De keuze van een zorgprofessional om een EPD volledig te gebruiken wordt EPD adoptie genoemd. Het doel van dit onderzoek is het inzichtelijk maken van de belemmeringen, oplossingen voor de belemmeringen, en de stimulerende factoren van EPD adoptie. Verder streven wij ernaar om te bepalen hoe adoptie modellen gebruikt worden bij het bepalen van een implementatie strategie en op welke manieren ziekenhuizen en patiënten voordelen ondervinden van een hoger niveau van EPD adoptie. Om dit doel te bereiken bevat deze thesis een literatuuronderzoek en een kwalitatief onderzoek, bestaand uit semigestructureerde interviews.

Het literatuuronderzoek bevat twintig geïncludeerde artikelen van PubMed, Web of Science, LISTA, Google Scholar, en HIMSS Analytics. In deze artikelen hebben wij belemmeringen, oplossingen voor belemmeringen, stimulerende factoren, en voordelen van EPD adoptie gevonden. We hebben dertig belemmeringen, zeven oplossingen, en negen stimulerende factoren gevonden, en we hebben deze gecategoriseerd in een taxonomie.

Met de interviews wilden we de volgende drie concepten vastleggen: EPD adoptie,

domeinen van het EPD, en het implementatie proces en de voordelen. De interviewer heeft elk interview opgenomen, uitgeschreven, en gecodeerd. We hebben achttien mensen geïnterviewd die betrokken waren bij een EPD implementatie in een Nederlands ziekenhuis, en daaruit hebben we vijfendertig belemmeringen, drie oplossingen, en zevenendertig stimulerende factoren voor EPD adoptie gevonden.

De belemmeringen voor EPD adoptie in zowel de literatuur als in de Nederlandse ziekenhuizen waren voornamelijk in de categorieën technisch, tijd, en sociaal. De oplossingen voor de belemmeringen waren voornamelijk in de categorieën technisch, psychologisch, juridisch, en veranderingsprocessen. De stimulerende factoren waren voornamelijk in de categorieën technisch en sociaal. In de interviews zijn twee

belemmeringen en twee stimulerende factoren gevonden, die niet voorkwamen in het literatuuronderzoek. Deze belemmeringen en stimulerende factoren zijn gerelateerd aan de meer geavanceerde aspecten van het EPD. Deze aspecten zijn een klinisch data warehouse, medicatie registratie, klinische beslissingsondersteuning, en informatie uitwisseling. Uit de interviews hebben wij geconcludeerd dat IT management zelden gebruik maakt van adoptie modellen in het bepalen van hun implementatie strategie. De voordelen die ziekenhuizen met een hoog niveau van EPD adoptie benoemen zijn een verbeterde efficiëntie van zorgprocessen, een verbeterde kwaliteit van zorg, en veiligheid van zorg. Als toevoeging op deze voordelen, gaven de geïnterviewde personen aan de patiënt meer controle te willen geven door middel van het gebruik van patiëntportalen. De bevindingen van deze studie kunnen gebruikt worden in het bereiken van de gevonden voordelen door het niveau van EPD adoptie te verhogen. Uit dit onderzoek blijkt dat er veel belemmeringen kunnen zijn die een hoger adoptie niveau in de weg staan. Deze studie beschrijft mogelijke oplossingen voor het overwinnen van deze belemmeringen en geeft een overzicht van bestaande

H.C. Vroege 3 | P a g e

Chapter 1

Introduction

Every year the costs of healthcare are increasing because of an increase in volume and quality, higher expectations, more guidance of patients, and because of an aging

population1. Hospitals are looking for ways to improve the quality of care while containing or even reducing costs. The use of IT in healthcare is seen as a method to both improve the quality of care and to contain or reduce costs, however this effect of IT in healthcare has not yet been proven 2. Hospitals in the Netherlands are improving their healthcare IT capabilities by implementing and expanding different aspects of their application environment, of which the central component is the Electronic Medical Record (EMR). The EMR is composed of the clinical data repository, clinical decision support, controlled medical vocabulary, order entry, computerized provider order entry, pharmacy, and clinical documentation applications 3. To improve the quality of care and contain the costs of care through the use of an EMR, the EMR has to be adopted by the healthcare organization. The adoption of an EMR is a complex process which involves many stakeholders, such as physicians and nurses of different

specialisms, ancillary departments, and administrative personnel. During the

implementation of an EMR the work of caregivers continues, which can lead to sudden changes in existing work processes. These changes require a reliable and usable EMR system and the caregivers are expected to cope with these changes. Because the path to a complete EMR is complex and has to be accepted by the entire organization, hospital organizations are looking for ways to strategize this path in order to achieve a higher adoption rate.

A possible tool to help hospitals in doing so is an adoption or maturity model. In Rogers’ theory of Diffusion and Innovation adoption is defined as: “a decision to make full use of an innovation at the best course of action available” 4. An adoption or maturity model describes how users are influenced in making this decision. These models can help improve the

adoption of an EMR by providing insight into what factors influence the adoption of the EMR.

Research objectives

The aim of this research is to find the barriers hospitals face and the facilitators which can help in reaching a higher rate of EMR adoption. Furthermore we aim to determine how an adoption model can help in achieving a higher adoption rate and in what way hospitals benefit from a higher adoption rate. Once the barriers and facilitators are known, possible solutions to overcome the identified barriers can be found.

The research questions used to reach our aim are:

1) What barriers prevent hospitals in the Netherlands from reaching an advanced stage of EMR adoption?

2) What solutions to these barriers exist?

3) What facilitators help hospitals in the Netherlands to reach an advanced stage of EMR adoption?

4) To what extent do adoption models help hospitals in reaching an advanced stage of EMR adoption?

5) What is the added value of an advanced stage of EMR adoption for hospitals and patients?

H.C. Vroege 4 | P a g e

The expected outcomes of this study are a list of barriers, facilitators and solutions to barriers from the literature and from people involved in EMR implementation in the Netherlands, a description of the perceived effects on quality of care and healthcare costs from hospitals that have achieved a high rate of adoption and expected benefits from those who are working towards higher adoption, and a description of the role of adoption models in this process.

Chapter organization

Chapter 2 will give relevant background information on the electronic medical record, EMR adoption, and existing adoption models. Chapter 3 reviews literature about barriers and facilitators to EMR adoption. Chapter 4 presents the methods and results of the interviews held with C(M)IO’s of hospitals. Chapter 5 presents the discussion of this research project, where we will touch upon strengths and weaknesses of the research methods, followed by recommendations for future research. Finally, we summarize the conclusions of this research project and provide answers to the research questions.

H.C. Vroege 5 | P a g e

Chapter 2

Adoption of the Electronic Medical Record

This chapter will explain what electronic medical record adoption is and which adoption models exist. First we will define what an electronic medical record is, than we will explain what adoption is, and finally we will give an overview of existing adoption or acceptance models.

Electronic Medical Record

More than one definition of the EMR exists. The electronic medical record as defined by HIMSS is: “ an application environment composed of the clinical data repository, clinical

decision support, controlled medical vocabulary, order entry, computerized provider order entry, pharmacy, and clinical documentation applications. This environment supports the patient’s electronic medical record across inpatient and outpatient environments, and is used by healthcare practitioners to document, monitor, and manage health care delivery within a care delivery organization (CDO). The data in the EMR is the legal record of what happened to the patient during the encounter at the CDO and is owned by the CDO” 3. The basis of the EMR is the clinical data repository (CDR), which is a real-time processing database of patient clinical information for practitioners. The clinical decision support (CDS) guides practitioners activities related to protocols and outcomes. The CDS will not perform as practitioners expect without a functional controlled medical vocabulary (CMV), which normalizes data from a relational and definitional hierarchy that enables other components of the EMR to optimally operate. Computerized provider order entry, pharmacy, and clinical

documentation applications are used to improve patient safety and reduce medical errors. A definition from Gartner for the Electronic Health Record (EHR) is: “An electronic health record (EHR) system contains patient-centric, electronically maintained information about an

individual’s health status and care, focuses on tasks and events directly related to patient care, and is optimized for use by clinicians. The EHR provides support for all activities and processes involved in the delivery of clinical care. The definition of an EHR system limits its scope to the continuum of care in one HDO.”5 This definition also explains that the EHR is used to maintain and process information about a patient’s health within one health delivery organization.

EMR Adoption

Because the success of an EMR is dependent on the users of the EMR, it is important to motivate and enable these users to use the EMR. The level of user acceptance of the EMR and the users’ ability to work with the EMR is what you could call EMR adoption. EMR adoption is faced with a vast amount of barriers, which will be described in the following chapters, which makes the process of EMR adoption a complex one.

In Rogers’ theory of Diffusion and Innovation adoption is defined as: “a decision to make full use of an innovation at the best course of action available” 4. Translated to the healthcare setting this could be described as: “A decision of a health care professional to make full use of the EMR at the best course of action available”.

H.C. Vroege 6 | P a g e

Adoption models

To help hospitals achieve a higher adoption rate different adoption models have been made. In the following section we will describe different variants of the Technology Acceptance Model (TAM), the EMR Adoption Model (EMRAM), and the Gartner generation model.

Technology Acceptance Models

The Technology Acceptance Model (TAM) is a model based on the Theory of Reasoned Action (TRA) of Ajzen and Fishbein’s. It was developed by F. Davis and R. Bagozzi. The TAM shows the relation between system design features, perceived usefulness and perceived ease of use, the users’ attitude toward using the system, and the users’ actual system use [Figure 1]. Davis states that the model can be used to “address why users accept or reject information technology and how user acceptance is influenced by system

characteristics” 6. This model was expanded into the TAM 27 and into the Unified Theory of Acceptance and Use of Technology (UTAUT) model8. The TAM 2 model splits the system design features into 5 different factors that can influence the perceived usefulness or the intention to use. These factors are: subjective norm, image, job relevance, output quality, and result demonstrability. The subjective norm can be influenced by the users’ experience and voluntariness [Figure 1]. The definitions of these influencing factors as stated by

Venkatesh can be found in Table 1 9.

Table 1 Influential factors TAM2 9

Factor Definition

Subjective norm Person's perception that most people who are important to him think he should or should not perform the behavior in question.

Image The degree to which use of an innovation is perceived to enhance

one's status in one's social system.

Job relevance Individual's perception regarding the degree to which the target system is relevant to his or her job.

Output quality The degree to which an individual believes that the system performs his or her job tasks well.

Result demonstrability Tangibility of the results of using the innovation.

The UTAUT model shows the influence of performance expectancy, effort expectancy, social influence, and facilitating conditions on the users’ behavioral intention and actual use behavior. The UTAUT models also takes into account the effect of gender, age, experience, and voluntariness of use on the influencing factors. All three of these models show that the users’ perception or expectations of the system influence the actual use of the system, which can be called adoption. Taking these factors into account during the implementation of an EMR can influence the adoption rate.

H.C. Vroege 7 | P a g e

Figure 1 TAM & TAM2

HIMSS Electronic Medical Record Adoption Model

Because a full electronic medical record consists of multiple components which can be implemented separately, the HIMSS developed an adoption model to help CDOs to assess their EMR capabilities and to strategize their path to a complete EMR. This adoption model is the HIMSS Electronic Medical Record Adoption Model (EMRAM). EMRAM consists of eight stages, starting with stage 0 (no EMR) and ending with stage 7 (complete EMR). The requirements of each stage consist of a subset of components required for the EMR, as is shown in figure 2. In stage 6 and stage 7 of EMRAM the most significant changes are made. For stage 6 these changes are the implementation of full physician documentation, a Clinical Decision Support System (CDSS) that guides all clinician activities related to protocols and outcomes, and a fully implemented closed loop medication administration environment. For stage 7 the hospital should no longer use paper to deliver and manage patient care and the patient chart should consist of a mixture of discrete data, document images, and medical images. Stage 7 also requires the hospital to use a Clinical Data Warehouse to analyze patterns of clinical data to improve quality of care and patient safety and to feed outcomes reports, Quality Assurance, and Business Intelligence. The final requirement for stage 7 is that clinical information can be readily shared via standardized electronic transactions with all entities who are authorized to treat the patient, or a health information exchange10 [figure 2]. HIMSS developed a methodology to provide peer comparisons for hospital organizations, where they use algorithms to automatically score hospitals in the database relative to the IT enabled clinical transformation status. If a hospital wants to reach stage 6, an onsite interview should be held and an online questionnaire is to be filled out. To reach stage 7 a more extensive inspection is carried out. This inspection consists of an all-day onsite visit of a HIMSS inspector, who checks all different components required for stage 7.

H.C. Vroege 8 | P a g e

Figure 2 HIMSS European EMR Adoption Model 10

In the Netherlands, 11 hospitals have reached stage 6 of the EMRAM. The EMRAM score distribution in Europe is shown in figure 3. The Netherlands is Europe’s EMRAM leader, with the highest percentage of stage 5 and 6 hospitals in the third quarter of 2014, 13% and 46.4% respectively. Even though this is the highest percentage of stage 5 and 6 hospitals in Europe, there still are numerous hospitals at a lower stage.

H.C. Vroege 9 | P a g e Gartner’s Generation Model

In 1998 Gartner made a five-stage model describing how EMRs would evolve. Gartner stated that EMRs would transform from very simple systems that provide results reporting into very complex, fully integrated systems that clinicians would use to help facilitate the practice of evidence-based medicine11 [Figure 4]. Table 2 shows the definitions for each of the

generations.

Table 2 Gartner’s Generations12

Generation Definition

The Collector Simple systems that provide a site-specific encounter-based solution to accessing clinical data.

The Documenter Basic systems that clinicians begin to use at the point of care for somewhat more than merely accessing clinical data.

The Helper More-advanced systems that support clinical episodes and encounters clinicians. The Colleague Advanced systems that provide substantial functionality for nurses, physicians and

pharmacists. These systems have more decision support and workflow capabilities. The Mentor Complex, sophisticated and fully integrated context-aware systems that cover the full

continuum of care and care givers, and that can actually guide clinicians when appropriate.

H.C. Vroege 10 | P a g e

Figure 4 Gartner's EMR Generations 11

The adoption models described here can help in formulating a strategy for EMR adoption and clarify the difficulties encountered in the adoption process. However, more specific knowledge of barriers and facilitators to EMR adoption is also needed, and will be investigated in the following chapters.

A Literature Review

H.C. Vroege 11 | P a g e

Table 3 Keywords used in literature search

Chapter 3

Barriers and Facilitators to EMR Adoption

A Literature Review

Introduction

EMR adoption can be defined as ‘A decision of a health care professional to make full use of the EMR at the best course of action available’. EMR adoption is discussed in detail in Chapter 2. According to the EMR adoption score distribution of HIMSS [Chapter 2, figure 2] the EMR adoption rate is quite low in Europe, where most hospitals are at stage 2 of the EMR adoption model. To improve the adoption rate of EMRs it is important to know the existing barriers and facilitators to this adoption. A barrier to adoption is a factor which will negatively affect the decision of a health care professional to make full use of the EMR at the best course of action available. A facilitator to adoption is a factor which positively affects this decision. A benefit of EMR adoption is a positive effect of the EMR on either quality of care or cost of care. Various research on barriers and facilitators to EMR adoption exists. Therefore, the aim of this literature study is to give an overview of the barriers, facilitators, and expected benefits of EMR adoption in hospitals found in existing literature. The research questions we try to answer with this literature study are the following:

1. What are the known barriers to EMR adoption? 2. What are the known facilitators to EMR adoption?

3. What are the (expected) benefits of a higher state of EMR adoption?

Methods

To create an overview of known barriers, facilitators, and expected benefits of EMR adoption, a literature search was

performed. We searched PubMed, Web of Science, LISTA, and Google Scholar using combinations of the following keywords: Electronic Medical Record (EMR), Electronic Health Record (EHR), Electronic Patient Record (EPR), Health Information Technology, Adoption Model, Adoption, Barriers, Facilitators, Benefits.

Table 3 shows an overview of the combinations of keywords used.

For systematic reviews identified during the

search, we performed a forward cited reference search in Web of Science. A forward cited reference search shows all studies citing a specific study.

Combination of keywords

Electronic Medical Record Adoption Model Electronic Medical Record Adoption

Electronic Health Record Adoption Model Electronic Health Record Adoption

Electronic Patient Record Adoption Model Electronic Patient Record Adoption

Electronic Medical Record Adoption Barriers Electronic Medical Record Adoption

Facilitators

Electronic Medical Record Adoption Benefits Electronic Medical Record Benefits

A Literature Review

H.C. Vroege 12 | P a g e

In order to select relevant studies the following inclusion criteria were applied in the selection process: 1) The title should contain at least one of the following words, or combination of words: Electronic medical record, electronic health record, barriers to acceptance, barriers to adoption, implementation of EHR. 2) The study aims to identify barriers and/or facilitators to the adoption of Health IT (EMR, EHR, EPR), the study aims to identify the benefits of the adoption of Health IT, or the study contains a taxonomy of barriers and facilitators. 3) All studies must be related to hospital care.

Further literature about EMR adoption was found via HIMSS Analytics. Literature in this database includes white papers and business cases of both research performed by HIMSS as well as research done by other organizations. The white papers and business cases were included to find results on achieved benefits through EMR adoption. These results were not widely available in scientific publications.

HIMSS Analytics performs research in different fields of healthcare IT, where one of the fields is EMR or EHR adoption. The business cases are made by hospitals which won the Davies Award. The Davies Award is a prize awarded to health care organizations who have utilized health information technology to substantially improve patient outcomes while achieving return on investment. To show evidence of the achieved improvements the healthcare organization must submit a case study showing their return on investment and clinical value. The white papers were assessed following the inclusion criteria as described above. The business case studies were included if effects of the EMR on quality and costs were mentioned in the executive summary.

Information extraction

From the articles we extracted barriers, solutions to barriers, facilitators, and expected benefits. We considered a factor a barrier if the article mentioned it as a barrier specifically, or if a negative influence of the factor on adoption was described. We considered a factor a facilitator if the article mentioned it as a facilitator, or if a positive influence of the factor on adoption was described. We classified a factor a solution to a barrier if it was mentioned as a step taken to counteract a barrier, rather than a real facilitator to adoption (e.g. “slow wireless connection” could be a barrier, and “upgrading the wireless system” would be a solution to this barrier). A barrier, solution, facilitator, or benefit was considered to be reported in more than one study if both studies described the same underlying concept (e.g. “inadequate or no IT support” and “lack of technical support” would be considered

synonymous). In cases where more detail was reported on a particular barrier, this was also extracted and summarized. If a study contained a taxonomy of barriers and facilitators we also extracted the taxonomy (described below).

For the extraction of benefits we looked at measurements of quality of care measured with outcome measures, or a reduction or savings in costs. Improved quality or cost reduction or savings were also called Return On Investment (ROI) in articles, where a distinction is made between ‘Hard ROI’ and ‘Soft ROI’. The benefits which can be directly measured in terms of financial gains are called “Hard ROI”. Benefits which are harder to measure in terms of financial gain, for example improvement patient safety and quality, are called “Soft ROI”13

A Literature Review

H.C. Vroege 13 | P a g e

Taxonomy

We planned to extract taxonomies of barriers from existing studies and to combine taxonomies if we found more than one. Taxonomies were combined by merging similar categories. This synthesis was performed by one researcher. Barriers found in the literature were organized using the combined taxonomy. Barriers not fitting into the taxonomy were put into an “other” category. Finally we categorized and summarized the barriers in the “other” category. One researcher evaluated the barriers, facilitators, and solutions to barriers extracted from the articles and classified them in the combined taxonomy. It was not necessary to have the exact same wording as was used in the taxonomy; rather, the categories of the taxonomy were considered to be concepts and the items extracted from the literature classified with the matching concept.

Results

The first search identified 1144 articles in the database search and 18 articles through other sources. These 18 articles were found in the following ways: cited reference search ( 2 articles) and white papers via HIMSS Analytics (5 white papers). We also found eleven business cases of Davies Award winning hospitals. Of these hospitals, ten are located in the U.S. and one hospital is located in Spain.

PRISMA 2009 Flow Diagram

Records identified through database searching (n = 1144 ) Scr e e n in g In cl u d e d El ig ib ili ty Id e n tifi cat ion

Additional records identified through other sources

(n = 18)

Records after duplicates removed (n = 1162 )

Records screened (n = 1162 )

Records excluded (n =1110 )

Full-text articles assessed for eligibility

(n = 52)

Full-text articles excluded, with reasons

(n = 32 )

Studies included in qualitative synthesis

A Literature Review

H.C. Vroege 14 | P a g e

Study characteristics and quality.

We found eleven articles about barriers and facilitators to adoption, nine articles about benefits of adoption, two white papers, seven business cases, and two taxonomies. A summary of the included literature can be found in table 4.

Table 4 Selected studies

Author Publication

year

Title Study design

Wilson, B.14 2007 The Value of Healthcare IT (HIT)

A Practical Approach to Discussing and Measuring the Benefits of HIT Investments

White Paper

England, I.15 2007 Executive management and IT

innovation in health: identifying the barriers to adoption Two-stage study - qualitative interview - quantitative survey HIMSS EHR Adoption and Sustainability Work Group16

2008 Key Issues in EHR Adoption and

Sustainability

White Paper

Boonstra, A.17 2010 Barriers to the acceptance of electronic medical records by

physicians from systematic review to taxonomy and interventions

Systematic review

Castillo, V.18 2010 A knowledge-based taxonomy of

critical factors for adopting

electronic health records systems by physicians: a systematic literature review

Systematic review

McGinn, C.19 2011 Comparison of user groups’

perspectives of barriers and facilitators to implementing electronic health records: a systematic review

Systematic review

Greiver, M. 20 2011 Implementation of electronic

medical records

Theory-informed qualitative study

Qualitative study The Advisory Board Company, HIMSS Analytics 21

2012 EMR Benefits and Benefit Realization Methods of Stage 6 and 7 Hospitals

Survey (qualitative study)

Struik 22 2014 The preferences of users of

electronic medical records in hospitals: quantifying the relative importance of barriers and

Discrete Choice Experiment

A Literature Review

H.C. Vroege 15 | P a g e

facilitators of an innovation Ingebrigtsen

23 2014 The impact of clinical leadership on

health information technology adoption: Systematic review

Systematic review

Boonstra, A. 24

2014 Implementing electronic health

records in hospitals: a systematic literature review

Systematic review

Ben-Zion 25 2014 Critical Success Factors for Adoption

of Electronic Health Record Systems: Literature Review and Prescriptive Analysis

Literature Review / Prescriptive Analysis

Briley, A.26 2014 Truman Medical Centers Davies

Award Enterprise Application

Business Case

Busick, H.27 2014 Lakeland HealthCare Core Case

Study

Business Case Van Daele,

D.28

2014 University of Iowa Health Care

Davies Award Enterprise Application

Business Case

Moncho, V.29 2014 Marina Salud Davies Award

Enterprise Application

Business Case

Davis, K30 2012 Mount Sinai Medical Center Core

Menu Item Return on Investment

Business Case

Menzies, C.31 2011 Children’s Medical Center Dallas

Core Case Study

Business Case Randeree, E.32 2007 Exploring Physician Adoption of EMRs:

A Multi-Case Analysis Multi-Case Analysis

Mitchell, M.33 2013 Texas Health Resources Core Case Study Business Case

Barriers and solutions

We found two taxonomies of barriers to EMR adoption through a systematic literature review. These taxonomies were developed by Boonstra and Castillo. Boonstra categorized barriers to EMR adoption into eight categories, namely: Financial, Technical, Time,

Psychological, Social, Legal, Organizational, and Change Process 17. Castillo categorized critical factors for adoption from sixty-eight studies into six main categories, namely: User attitude towards information systems, Workflow impact, Interoperability, Technical support, Communication among users, and Expert support 18. To organize the barriers found in literature we combined the taxonomies of Boonstra and Castillo. Table 5 shows the overlap in categories of Boonstra and Castillo.

Table 5 Boonstra and Castillo’s taxonomy overlap

Boonstra’s categories Castillo’s categories

Financial

Technical Technical support

Interoperability

User attitude towards information systems

A Literature Review

H.C. Vroege 16 | P a g e

User attitude towards information systems

Psychological User attitude towards information systems

Social Communication among users

Expert support Legal

Organizational

Change process User attitude towards information systems

Although the Castillo taxonomy provided more detail than the taxonomy of Boonstra, Castillo’s taxonomy was a subset of that of Boonstra. Thus, only the high-level categories from the Boonstra taxonomy remained in the final combined taxonomy that was used in subsequent analyses.

We found thirty distinct barriers, six distinct solutions to barriers, and nine distinct facilitators in 8 categories, with one to seven barriers/facilitators per category. Table 6 shows an overview of the barriers found in each study and table 7 shows an overview of the facilitators found in each study. In the following part we will describe the barriers, solutions to barriers, and facilitators, of each category in more detail.

Table 6 Barriers to EMR adoption

Category Barrier Found in study

Financial High start-up costs 17, 19, 16, 25

High Ongoing costs 17, 19, 16, 25

Uncertainty over ROI 17, 19, 15, 25

Lack of financial resources 17, 19,24

Technical Lack of computer skills 17,18, 19, 24

Lack of training and support 17, 18, 19, 22,25 Complexity of the system 17, 18, 19, 15,20, 25

Lack of customizability 17, 18,19,22

Lack of reliability 17, 19, 24

Interconnectivity / standardization 17, 18, 19, 24, 15, 25

Lack of hardware 17, 24, 22

Time Time required for the implementation process 17, 18, 19, 24, 15

Time to learn the system 17, 18, 19, 15

Time required to enter data 17, 18, 19

More time per patient 17, 18, 19

Time to convert patient records 17, 24, 15

Psychological Lack of beliefs in EMRs 17, 18, 24, 15, 25

Need for control 17,25

Social Uncertainty about the vendor 17, 24

Lack of support from external parties 17, 18, 24 Interference with doctor-patient relationship 17, 18, 19, 25 Lack of support from other colleagues 17, 18, 24 Lack of support from the management level 15,17,22,24

Legal Privacy or security concerns 17, 19, 24, 25

A Literature Review

H.C. Vroege 17 | P a g e

Organizational type 17, 24, 23

Change process Lack of support from the organization culture 17, 18, 24, 25

Lack of incentives 17

Lack of participation 17, 24

Lack of leadership 15,17,22–24

Financial (four barriers)

High start-up costs + high ongoing costs (four studies): All four studies reported that the financial burden of the high initial and ongoing costs of an EMR can be a large barrier to EMR adoption for hospitals 16,17,19,25. The white paper “Key issues in EHR adoption” from HIMSS showed that in 2008 initial capital costs related to the acquisition of EMR applications have been estimated to range between $15,000 and $80,000 per physician16. Ongoing costs can range between 20-30% of the acquisition costs of the system per year16.

Uncertainty over ROI (four studies): All four studies reported that uncertainty over ROI exists because there is a lack of evidence on the impact of IT on health care 15,17,19,25. Business cases of Davies Award winning hospitals show that a ROI becomes visible after approximately five years 26–29,33.

Lack of financial resources (three studies): Two studies mentioned this barrier in

combination with the high start-up costs and high ongoing costs 17,19. The lack of financial resources was especially found in the smaller hospitals, who have a lower budget available than the academic and top-clinical hospitals 17,24.

Technical (seven barriers, three solutions to barriers)

Lack of computer skills (four studies): All four studies reported that lack of computer skills can lead to physicians resistance to EMRs because physicians fear that they are not capable enough to use the system, which will reduce their performance 17–19,24. Two studies mention a specific computer skill that physicians can lack is typing skills, which is needed to enter patient data. The lack of this skill can result in medical errors because of typing errors 17,24. A proposed solution to this technical barrier is the availability of technical support17.

Lack of training and support (five studies): All five studies reported that due to the lack of computer skills of end-users and the complexity of an EMR, end-users require training and support 17–19,22,25. Boonstra reported that without training and support being available, the end-user will be reluctant to use the EMR 17. A solution to this technical barrier was also the availability of technical support17. Castillo and Struik state that technical support can take away some of the uncertainty of the end-users of EMR systems and therefore contribute to the adoption of the EMR 18,22. Options for technical support mentioned by Castillo are support via social interactions, for example helpdesks, or by documentation18. Another proposed solution was giving professional training to the users, or by support offered by more skilled colleagues17–19,24.

Complexity of the system (six studies): All studies reported that users have to invest time and effort to master a complex system, which slows the adoption15,17–20,25. McGinn describes a complex system as a system which has confusing screens, options, and navigational aids19. Boonstra mentions a possible relation between the complexity of the system and the lack of computer skills of a user, as a less skilled user might perceive the system as more complex 17. Lack of customizability (four studies): Randeree describes customizability as the ability to be adapted of the technology system that fails to conform to specific needs of the user

A Literature Review

H.C. Vroege 18 | P a g e

applications 32. Two variants of this barriers are found in the studies. First, Boonstra reported a lack of customizability as a situation where no suitable application can be found that meets the users’ needs 17. Three studies reported the second variant, which occurs when the selected application cannot be customized to the users’ preferences 18,19,22.

Lack of reliability (three studies): Boonstra defines reliability as “the dependability of the technology systems that comprise the EMRs”17. All three studies report that if high reliability cannot be reached users will be reluctant to use the system17,19,24. A possible solution to this barrier mentioned by Boonstra is to promote and communicate the reliability and availability of the system17.

Interconnectivity and standardization (six studies): All six studies reported compatibility problems as a barrier to adoption17, 18, 19, 24, 15, 25. Before an EMR can be used as one

integrated system, multiple applications must be interconnected with each other. The same holds for the exchange of data where the data can hold different formats for different applications, which is a lack of standardization 17, 18, 19, 24, 15, 25.

Lack of hardware (three studies): All three studies reported that every EMR system needs a basic set of hardware and for higher levels of maturity more hardware might be needed. If the required hardware is not available the adoption of an EMR is not possible17,22,24. This barrier can also lead to a financial barrier, as the purchase of hardware can be too expensive 17,22,24

.

Time (five barriers)

Time required for the implementation process (five studies): All five studies reported that physicians would rather not invest time in the implementation of an EMR, because they should spent this time on treating patients15,17–19,24.

Time required to learn the system (four studies): All four studies reported that physicians feel that the time that they need to learn the system reduces their productivity17, 18, 19, 15. Time required to enter data (three studies): Three studies reported time required to enter data as a barrier to adoption17–19. Boonstra mentions that this barrier is related to the complexity of the system, or the computer skills of the user17.

More time per patient (three studies): Three studies reported that more time is required per patient, which is a barrier17–19 Two studies report that this barrier is related to the previously mentioned barrier, because the increase in time required to enter data reduces the time available during a consult17,19.

Time required to convert patient records (three studies): Three studies report the time required to convert patient records as a barrier15,17,24. When a hospital switches from a paper-based record to an EMR, the transfer of records between systems takes time in which this records can be unavailable15,17. Boonstra mentions that some physicians consider this transfer a personal responsibility17.

Psychological (two barriers, two solutions)

Lack of belief in EMRs (five studies): Five studies reported the lack of belief in EMRs as a barrier. This lack of belief means that end-users doubt that an EMR can improve patient care or clinical outcomes. When the users do not see the benefits of an EMR they are reluctant to use it, which hinders the adoption17, 18, 24, 15, 25. A solution to this barrier reported by

Boonstra and Castillo is that fellow physicians demonstrate the ease of use and usefulness of the EMR 17,18. A solution stated by Ben-Zion is that it can be helpful to show direct benefits

A Literature Review

H.C. Vroege 19 | P a g e

to physicians, such as greater efficiencies in administration, governance, research, and patient care25.

Need for control (two studies): Boonstra and Ben-Zion report this barrier as the perception of the effect an EMR has on the professional autonomy of physicians17,25. Boonstra states that physicians fear that the implementation of an EMR will reduce their control of patient information and working processes can exists, which forms a barrier to adoption 17. Ben-Zion describes physicians as powerful actors in healthcare organizations who might find it difficult to force a change in work patterns25.

Social (five barriers)

Boonstra describes social barriers as a flaw in a relationship between medical practitioners that have to make EMR decisions and other parties in healthcare industry, such as vendors, subsidizers, insurance companies, patients, administrative staff, and managers 17. The social barriers found in the literature are uncertainty about vendors, lack of support from external parties, interference with doctor-patient relationship, lack of support from other colleagues, and lack of support from the management level.

Uncertainty about vendors (two studies): Both studies reported uncertainty about vendors as a barrier which can arise when the hospital has problems with selecting a vendor. Both studies state that the vendor should be able to offer a mature and successful product and the vendor must be able to identify hospital workflows and adapt its product to these workflows 17,24.

Lack of support from external parties (three studies): Three studies reported that the lack of support by external parties, such as government or insurance companies, can result in decisions that oppose EMR adoption 17,18,24.

Interference with doctor-patient relationship (four studies): Four studies reported an interference in communication between physicians and patients as a barrier. All four studies reported this barrier in combination with lack computer skills, which increases the time a physician spends on entering data instead of listening to the patient 17–19,25. McGinn found that patients had concerns about receiving bad news electronically instead of in person 19. Lack of support from other colleagues(three studies): Three studies reported lack of support from colleagues as a barrier when employees are not willing or capable to use the EMR applications. This lack of support from colleagues can impede physicians to further adopt the EMR 17,18,24.

Lack of support from the management level (four studies): Struik22 reported that

managerial support can be divided into two different levels, biding and stimulating. On the biding level the head of department should emphasize that the use of the EMR should interfere the regular work as little as possible. On the stimulating level the head of

department should emphasize the importance of good use of the EMR for the quality of the work of the department. The stimulating attitude of the head of department leads to more acceptance of the EMR 22. All four studies reported that the lack of management support will not contribute to more acceptance and thus slow the EMR adoption 15,17,22,24. England

reported multiple reasons for the lack of management support, namely that managers can: lack confidence in the available IT solutions, face many conflicting demands for resources, find IT vendors act inappropriately, do not believe there is a compelling business case for IT investment, or do not believe that effective clinical IT exists15.

A Literature Review

H.C. Vroege 20 | P a g e

Legal (one barrier, one solution)

Privacy or security concerns (four studies): Four studies reported privacy or security

concerns for security breaches and the safety of information storage17,18,24,25. Ben-Zion states that a main challenge for health care organizations is cloud computing. This is associated with security and privacy because medical records are highly sensitive and should be

protected according to existing regulations 25. A solution reported by Boonstra to this barrier is to develop the security requirements in cooperation with physicians and patients and to ensure that the EMR meets the requirements before implementation17.

Organizational (two barriers)

The size of the organization (one study): Boonstra reported that the size of an organization is important in both the selection procedure and the EMR adoption after selection. Larger hospitals usually require more time for the selection and implementation process than smaller hospitals, but the adoption rate in larger hospitals is higher17. Boonstra reported that a reason for this higher adoption rate is that larger organizations have more extensive

support and training systems. This is also related to the financial situation of the hospitals, where larger hospitals usually have a larger budget for IT investment 17.

Organizational type (three studies): Three studies reported that the organizational type can slow the process of adoption, for example in a highly bureaucratic organizational

structure17,23,24.

Change process (four barriers, one solution)

Lack of support from the organizational culture (four studies): Ben-Zion reported that the organizational culture should be one of innovation, exploration, and continual improvement, where both positive and negative impacts of EHR use are communicated openly25. All studies reported that if such a culture is not present in a health care organization this can be a barrier to EMR adoption, because the organization will not be supportive of the new EMR17,18,24,25.

Lack of incentives (one study): Boonstra reported that physicians need to see some personal benefit from using EMRs, otherwise they will not be motivated to change their working procedures17.

Lack of participation (two studies): Two studies reported lack of participation as a barrier to adoption17,24. A solution to this barrier mentioned in four studies16,17,23,25 is the introduction of a key user, which is preferably an experienced physician. The role of the key user is that of a project champion who can provide support for his colleagues and communicate the

advantages for physicians to other colleagues16,17,23,25.

Lack of leadership (five studies): Five studies reported that leadership is required to motivate the users and stimulate the adoption process15,17,22–24. Boonstra reported that without project leaders hospitals may struggle to improve quality or see financial benefits from EMRs17, and that an important aspect of leadership is the communication of a comprehensive implementation strategy24.

A Literature Review

H.C. Vroege 21 | P a g e

Facilitators to EMR adoption

Nine facilitators were found in three studies. Table 7 shows an overview of these facilitators.

Table 7 Facilitators to EMR adoption

Financial (one facilitator)

IT funding (one study): Ben-Zion reported IT funding provided by the government or insurance companies as a facilitator to adoption25.

Technical (four facilitators)

Adoption of standards (two studies): Castillo and Ben-Zion reported that the adoption of standards is important for an EMR to be adopted by end-users18,25. Castillo mentions that documentation about the used standards should be available for all users in order to improve the use of standards18.

Accessibility and compatibility (two studies): Two studies reported compatibility of systems with each other and compatibility of systems with the workflow to be a facilitator to

adoption18,25. Ben-Zion specifically mentions the use of mobile devices to improve accessibility25.

Friendly user interfaces (one study): Castillo reported that friendly user interfaces influence the users’ attitude towards the EMR18.

Access to local FAQ lists (one study): Castillo reported that the availability of FAQ lists help users to learn the system more easily 18.

Time (one facilitator)

Overall increased productivity in the overall process of patient care (one study): McGinn reported that from a management and patient perspective the overall process of patient care was perceived as more efficient after the implementation of an EMR 19. McGinn also reported that physicians see a decrease in their personal productivity19.

Social (three facilitators)

Electronic bulletin boards (one study): Castillo reported that electronic bulletin boards can be used to communicate both positive and negative remarks about the use of the EMR18. Partnership with companies (one study): Ben-Zion reported that partnerships with healthcare innovation companies, consulting firms, and industry peers can lead to more IT innovation which meets the user requirements25.

Category Facilitator Found in study

Financial IT funding 25

Technical Adoption of standards 18,25

Accessibility and compatibility 18,25

Friendly user interfaces 18

Access to local FAQ lists 18

Time Overall increased productivity 19

Social Electronic bulletin boards 18

Partnership with companies 25

A Literature Review

H.C. Vroege 22 | P a g e

Collaboration between the clinical staff and IT executives (one study): Ben-Zion reported that collaboration between the clinical staff and IT executives will result in a more user friendly EMR25.

Benefits of EMR adoption

Benefits were not described in any of the scientific literature found in this study. All seven Davies Award case studies describe the benefits realized from EMR adoption in terms of both hard ROI and soft ROI, and, in the American hospitals, a distinction is made between ROI from the Meaningful Uses program and ROI from their own investments. Of the available Davies Award case studies, six are from American hospitals and one is from a Spanish hospital. Table 8 shows the sources of ROI found in these hospitals. One white paper and one survey study performed by HIMSS also investigated the benefits of EMR adoption.

Hard Return on Investment

Nine sources for hard ROI were found. These sources are improved employee productivity, reduction of average costs of care, reduction of transcription expense, sunset application savings, clinic and HIM chart control staff reduction, lab test utilization, monitoring device integration, reduction in FTE’s, and real estate optimization.

According to Wilson the overtime expenditure per 1000 admissions was lowered by 5.3%. ROI was also achieved by reducing the average cost of care. A way of reducing these costs are the reduction in ADEs. Wilsons’ research shows that ADE cost avoidance per 1000 acute admissions is increased by 84.3% because of the use of health IT 14. Another hard ROI source is the sunset application savings, which means that hospitals shut down redundant or

obsolete business applications while retaining access to the historical data. Removing these applications allows IT departments to reduce the software, hardware, and resources

required to manage legacy data.

Table 8 ROI benefit sources26–31

Hard ROI – Benefit source Soft ROI - Benefit source

Improved employee productivity Reduced medication ADEs

Reduce average cost of care Reduced pressure ulcers

Reduced transcription expense Radiation safety program Sunset application savings Reduction in patient falls Clinic & HIM chart control staff reduction Reduction in CAUTIs

Lab test utilization Reduction in colorectal surgical infections

Monitoring device integration FTE’s reduction

Real estate optimization

Soft Return on Investment

Six sources for soft ROI were found. These sources are reduced medication Adverse Drug Events (ADEs), reduced pressure ulcers, radiation safety program, reduction in patient falls, reduction in CAUTIs, and a reduction in colorectal surgical infections. All of these sources are an increase in quality and also lead to a reduction in the average cost of care, which is a part of hard ROI. A qualitative study performed by the HIMSS advisory board showed that

hospitals achieved benefits in quality for multiple core measures, such as VTE, congestive heart failure, and strokes 21.

A Literature Review

H.C. Vroege 23 | P a g e

Discussion

The aim of this literature study was to give an overview of the barriers, facilitators, and expected benefits of EMR adoption found in existing literature. In this study we classified thirty barriers to EMR adoption into eight categories defined in the combined taxonomy of Boonstra and Castillo. The majority of barriers identified from the literature fell into the categories Technical (seven barriers in eight studies), Time (five barriers in five studies), Social (five barriers in seven studies), and Change process (four barriers in seven studies). Relations between barriers exists, where barriers can increase the effect of other barriers, or where barriers can be the cause of other barriers. We found seven solutions to barriers in the categories Technical (two solutions in five studies), Psychological (two solutions in three studies), Legal (two solutions in one study), and Change process (one solution in five

studies). We found nine facilitators in the categories financial (one facilitator in one study), technical (four facilitators in two studies), time (one facilitator in one study), and social (three facilitators in two studies). In total fifteen sources of Return on Investment were found, of which nine were hard ROI and six were soft ROI. Hard ROI was achieved through cost reduction, increased productivity, and savings on IT maintenance. Soft ROI was achieved through a quality increase for multiple core measures.

A limitation of this study is that the categorization was performed by only one researcher. Although we combined two taxonomies from the literature, the use of a different taxonomy, or if the categorization was done by a different or more researchers, could have let to

different results. However, the taxonomy of Boonstra uses high-level categories of which the taxonomy of Castillo was a subset, and other taxonomies might be as well. Although we employed an extensive search strategy using several databases, it is always possible that some relevant studies were missed. However, research of barriers to the adoption of EMRs has been done extensively and the results of these studies were quite similar. Therefore we can say that the found barriers probably give a fairly complete overview of existing barriers to EMR adoption. However, not much research was found on the benefits of EMR adoption. The benefits were almost all derived from the business cases performed for Davies Award winning hospitals in America. Therefore the positive results might not be applicable to European hospitals. The American hospitals also benefitted from the Meaningful Use (MU) program. Through this program hospitals could receive funds for the purchase of an EMR provided that they satisfy the conditions of the program. The main goal of these

measurements is that hospitals can show that their EHR is used in a way that can positively affect the care of their patients34. These hospitals did separate the Meaningful Use benefits from the benefits achieved from their own investments, but it is hard to quantify the real effect of the Meaningful Use program. A final limitation to this study is the inclusion of systematic reviews. The use of systematic reviews could introduce a selection bias because the quality and completeness of our results are partially dependent on the quality of the methods used in these studies. However, our aim was to provide an overview of existing barriers and facilitators, which has been done in the systematic reviews. Because the results of these studies are quite similar we can say that the risk of bias introduced by the use of systematic reviews is limited for our study. The systematic reviews included in this study covered either barriers to EMR adoption17,19,23, facilitators19,24, or critical (success) factors18,24 for adoption. This study added a review of the benefits of EMR adoption and gave an

H.C. Vroege 24 | P a g e

On the basis of this literature study, we can conclude that the barriers found in the literature were mainly in the technical, time, and social categories. The solutions to barriers found in the literature were in the technical, legal, psychological, and change process categories. The facilitators found in the literature were mainly in the technical and social categories. Benefits achieved through the implementation of the EMR were mainly an increase in efficiency and quality of care. The increase in quality was visible in a reduction in medical errors such as Adverse Drug Events. Our study adds an overview of benefits of EMR adoption, which was not addressed in previous systematic reviews. The literature study told us what barriers, facilitators, and solutions to barriers are known and what benefits can be achieved through EMR adoption, but we were still lacking information on barriers, facilitators, solutions to barriers, and benefits in the Netherlands. Therefore we conducted a study onbarriers, facilitators, and (expected) benefits of EMR adoption in the Netherlands, described in chapter 4.

Chapter 4

EMR Adoption in Dutch Hospitals

Introduction

In the literature review in the previous chapter, we learned that there are many barriers and facilitators to EMR adoption. However, we anticipated that the Dutch situation may be different than the situation described in the literature. Also the literature provided limited information on the benefits of EMR adoption. We sought to conduct semi-structured interviews with IT-management of Dutch hospitals to assess the state of EMR adoption in the Netherlands, to see if the barriers and facilitators in the literature exist in Dutch hospitals and to investigate the role of adoption models in EMR adoption.

Methods

Interview questions

With the interviews we wanted to capture three concepts related to EMR adoption. These concepts were: EMR adoption, domains of the EMR, and the implementation process and benefits. To determine what EMR domains exist we used the EMR definitions given in chapter 2 and the components listed in the HIMSS EMRAM10. Based on these a draft questionnaire was developed by one researcher (HV), and then the questions were

generalized to make them applicable to hospitals using any or no adoption model. The draft questionnaires were reviewed by a second researcher with experience in constructing interview questions (SM). To test the completeness and clarity of the interview questions we conducted a pilot interview.

Interviews and coding of results

We obtained the contact details of possible interviewees through an available contact list of a consultancy firm. We sent an email to every interviewee explaining the goal and setting of the interview and asked if they would like to participate. Each interview was recorded with permission of the interviewee and the interviewer took notes during the interview. The interviewer transcribed the recording of each interview verbatim (only the used wordings are transcribed, without non-verbal context) and coded each transcription. We used the same underlying concepts used to develop the interview questions to code the interview transcripts.

H.C. Vroege 25 | P a g e

The codes were: EMR Adoption

 Reasons for EMR implementation

 Digital status before new EMR

 Use of adoption models Domains of the EMR

 Domain + status (is it implemented)

 Domain + barrier

 Domain + solution to barrier

 Domain + facilitator

Implementation process and benefits

 Implementation process

 Implementation process: adoption barrier

 Implementation process: adoption facilitator

 Benefits hospital

 Benefits patient

If results did not fit this coding system, new codes would be added to categorize these results. After we coded the results we tried to classify the barriers and facilitators found in the interviews using the taxonomy used in the previous chapter. We put the barriers and facilitators that did not fit this taxonomy in an “other” category. We count the results in each category and categorize and summarize the results from the “other” category.The

interviews are all conducted, transcribed, and coded by the same interviewer.

Finally we compared the specific barriers and facilitators found in the interview results with the results of the literature review to determine if any new barriers are found.

To prevent bias toward ideas from the first interviews, we took notes of each comment that was classified into a category and compared the comments of all interviews in the end to determine the overall category. Within the results we made a distinction between academic and non-academic hospitals. This distinction is made because of the difference in care processes and the difference in available resources.

Results

Interview questions

The interview questions were formulated around three concepts, namely EMR adoption, Domains of the EMR, and Implementation process and benefits. These concepts each contain multiple questions, which can be found in Appendix B.

On the basis of the pilot interview, we made minor revisions to the interview questions and added one question. The added question was “What would you change if you could redo the implementation process?”.

Interview results

We contacted thirty possible interviewees of which four did not want to participate in an interview and eight did not respond. Eighteen people involved in an EMR implementation in a Dutch hospital (seventeen hospitals in total) were interviewed on the subject of EMR adoption. Of the eighteen interviewees, seven were Chief Information Officers (CIO), two were Chief Medical Information Officers (CMIO), five were Information Managers (IM), and