To answer the research question, a qualitative research study was performed. A qualitative design was considered appropriate for this question, as the primary objective was to explore more in-depth perceptions of factors and processes related to a more complex system, including social and technical components [13].

The development of EMR use in 73 Dutch hospitals in the period of 2012-2015 was measured using the EMRAM score [14]. The hospitals that were measured twice in the research period and did not work with nursing documentation in EMRs or with the CLMA and advanced CDSS were asked to participate in a follow-up study (Table 1). Two hospitals achieved a higher stage on the EMRAM score, eight hospitals stayed on the same level, and one did not respond (Table 1).

The hospitals were approached through the chairs of their medical staff and were asked to nominate a medical specialist each to participate in this study. To be eligible, the medical specialists were required to have worked for more than 5 years in the hospital in question. Overall, the selection of participants was geared toward a balanced mix of different specialties (surgical, nonsurgical, small specialty).

In the period of August-October 2018, a semistructured interview of about 90 min was conducted with each medical specialist selected. The abovementioned research question was at the core of the interview. An item list (available from the first author [RvP]) was used by the interviewer to help the participants focus on relevant experiences in case the conversation halted. We only asked questions about aspects the medical specialists were personally dealing with; we were not interested in second-hand accounts.

For analysis of the answers, we used the classification of aspects that can influence the implementation of EMR systems based on the taxonomy of Boonstra et al [15]. This systematic literature review was carried out to identify all the barriers that result in physicians showing resistance toward EMR systems. Table 2 shows the highlights of this taxonomy model.

We used this taxonomy for the same aspects in a neutral connotation, as the same taxonomy can be followed when categorizing aspects (quotes) as facilitators of EMR use. All authors participated in the allocation of quotes from the interviews to a category of the described taxonomy, with initial allocation by the first author (RvP) and validation by the second and third authors (AJPS and AB). The results were recorded in a Microsoft Excel file, which is available upon request from the first author (RvP). Based on this classification and the primary interview recordings, the authors reached a consensus about the best way to allocate quotes to the categories of the taxonomy. All authors agreed with the final allocation of quotes.

As this study did not involve research on human subjects, no medical ethical committee approval was required under Dutch law. Neither the Dutch Medical Research Involving Human Subjects Act (Wet Medisch-Wetenschappelijk Onderzoek met Mensen [WMO]) nor the university required ethics approval for the type of work conducted in this research. All participants orally and voluntarily agreed to participate in this study. They allowed us to use the data they provided, including quotes, under the condition of confidentiality. All participants agreed with the final report of their interviews. All participating hospitals, medical specialists, and used quotes in the manuscript were anonymized. No written permission was needed in this case.

In all, 11 hospitals (Table 3) agreed to participate in this qualitative study, while 3 hospitals (1 regional and 2 teaching hospitals) were unwilling or unable to participate. Each participating hospital nominated 1 medical specialist to be interviewed, representing 10 different specialties and between 5 and 27 years of experience in their current hospital. Six of them were (former) chairs of medical staff.

In total, the participants made 160 observations regarding aspects that influence the relationship between the extent of EMR use and the quality of care: 122 observations were characterized as barriers and 38 as facilitators. First, we will discuss the technical aspects that are mentioned most often. Next, we will discuss the other aspects of EMRs that influence the quality of care. Not every aspect of the taxonomy was used, because some aspects were not mentioned during the interviews. The legal aspect, related to information safety, was not mentioned in any of the interviews. The organizational aspect (type and size of the hospital) could not be addressed in the analysis because it was not part of the design of the study and was out of scope for individual participants. As explained in the Methods section, the change process aspect is of particular interest from a systems perspective, and it will be treated last to reflect on possible future developments.

The participating hospitals included two academic hospitals, four teaching hospitals, and five local hospitals (see also Tables 1 and 3). However, based on hospital type, no difference was found between participants’ observations. The availability of nursing documentation and of the CLMA and advanced CDSS in the hospital concerned did not lead to differences (see also Tables 1 and 3) in the experiences of the medical specialists.

During the interviews, technical aspects were the category all medical specialists chose to talk about first. It therefore seems like medical specialists consider technical aspects the most important factor influencing EMR use in hospitals. To provide more insight, the related tables include a subdivision of technical aspects, followed by quotes from participants to illustrate what the more abstract terms of the model mean. A complete list of all quotes is available from the first author (RvP).

In Table 7, other aspects of EMRs are summarized. These aspects were mentioned less often by the medical specialists but are also important influencing factors. These observations are generally consistent with the findings published elsewhere [17,18]. One thing that stands out is that medical specialists sometimes miss informal contacts of meetings with colleagues that were previously necessary due to the lack of a common digital file. Equally striking is that the financial aspect was hardly mentioned. The latter contrasts with findings in other publications [16].

In Table 8, illustrative quotes for the change process aspect are summarized. According to the medical specialists, several preconditions for success must be met before the successful introduction of EMRs in their hospitals. There is some doubt, for example, as to whether the supplier of EMRs is willing to create links to other parts of the IT system. However, this runs counter to market forces. Moreover, the participants mentioned that governmental institutions often also still require medical specialists to use paper. A central theme for almost all interviewed medical specialists is the coded or noncoded recording of obtained information. They generally realize that encoding a medical record is a prerequisite for getting help from the EMR system based on so-called evidence-based material. Several hospitals initially started out with the recording of this information by medical specialists but later abolished this system because the medical specialists refused to work with it.

To answer the research question “Which positive or negative aspects influence the relationship between EMR use and the quality of medical care according to medical specialists?”, a qualitative research study was performed.

The overall picture of the relationship between the extent of EMR use and the quality of medical care according to the participants shows that medical specialists prefer digital records over old paper ones. However, at the same time, the participants consider the technical systems old-fashioned compared to the systems they can access at home to book a trip or buy a book. The inability of all those involved (professional groups, boards, suppliers, politicians) to improve this situation is described openly by some participants. By and large, the participants do see the potential, but a better way to record coded information still needs to be found. The lack of interconnection between the different EMRs, for example, per hospital department (eg, internal medicine and cardiology), is also seen as an important limitation. Noteworthy is also that financial aspects are not mentioned often. This contrasts with other studies, in which technical issues and financial issues are mentioned in equal measure [12]. The obvious question is then whether money plays an important role according to medical specialists. Finally, these systems should be able to offer support in decision making for diagnosis and treatment [19].

The participants indicated that it is necessary to fulfil some of the preconditions for success before the EMR can make a positive contribution to the hospital’s daily practice. That is essentially the source of the medical specialist’s resistance. It takes a lot of effort and time to keep the patient file up to date. The only result is a nonpaper file, which medical specialists appreciate but is ultimately not enough to motivate them. It is tempting to make encoding medical data mandatory. However, without interventions in the organization, this is doomed to fail because many medical specialists are unwilling or unable to comply. In the end, inefficiently organized processes will then be automated at great cost and effort, while remaining inefficient at their core [20].

So, prior to the question of how to improve the available processes comes the basic question, What can be improved? Are the available business processes principally accepted, or can the search be directed toward a change in the existing processes [21]?

As stated, it is essential to understand what medical specialists perceive to be key aspects that either support or hinder the use of EMRs to positively impact diagnosis and treatment, now and in the future. These findings may help decide how medical processes can be improved with modern IT. Important in this approach is that the possibilities of modern IT, especially for advanced decision support, be taken as a starting point [22].

Our study had several limitations. We only interviewed medical specialists from hospitals that lacked nursing documentation in 2012-2014. These hospitals found themselves at the lowest stage of EMR use according to the EMRAM model but also had great potential to improve effectively and were able to learn from other hospitals. However, this qualitative study might still have been too early in the hospitals’ implementation of EMRs to identify aspects that are relevant for mature use of EMRs.

The age of the interviewed medical specialists varied between 42 and 62 years. These are medical specialists with a great deal of experience in the field. A question is whether the perspectives of younger medical specialists correspond with the perspectives of their older colleagues. Follow-up research might answer this question.

The 11 medical specialists shared positive experiences of EMR use when comparing it to the formerly used paper records. The fact that the health professionals involved can access patient data at any time they need it is considered important. However, in practice, potential quality improvement lags as long as decision support cannot be applied due to the lack of a fully coded patient record.