• No results found

Improving the communication between the hospital and general practitioners using health infromation technology

N/A
N/A
Protected

Academic year: 2021

Share "Improving the communication between the hospital and general practitioners using health infromation technology"

Copied!
68
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

the hospital and general practitioners

using health information technology

Uchaindani Maresa Watson

Thesis Master Medical informatics

(2)

Improving the communication between the hospital and

general practitioners using health information technology

Student

Uchaindani Watson

Student number: 10465863

Mentor

Dr. A. (Anniek) van Vlijm-Kiewiet Department of Internal Medicine Amsterdam UMC, Location AMC

Tutor

S.K. (Stephanie) Medlock, DVM, PhD Assistant Professor

Department of Medical Informatics, AMC-UvA

Location of Scientific Research Project

Department of Medical Informatics, Amsterdam UMC location AMC Department of Internal Medicine, Amsterdam UMC location AMC

Practice teaching period November 2017 – May 2019

(3)

Acknowledgements

I am grateful that I made it!!!

At the start of the Master thesis and during this project I went through several difficult times in my life. These were times that made me less motivated in life but also less motivated to finish this project. I also could not believe I was able to finish this project... However, I am thankful for the best tutor and mentor I could have ever asked for to guide me through this project: Stephanie (Ace) Medlock and Anniek Vlijm-Kiewiet. I want to thank you for your constructive feedback, guidance, patience and understanding.

I also would like to thank my cousin Latoya, Patty, my grandma, my mom, my dad, my sisters and brothers for their support. I thank the leaders and my friends from de Rots who all had a significant role in my life during this master thesis project. Special thanks to Elfriede, Rochee, Sharon, Hans, Marlon, Rolito, Sander and Letitia. I’m grateful for this journey and I am thankful for the students that have become friends for life (Jolien, Priya, Safira, Shuxin and Kenneth).

(4)

Content

Acknowledgements ... 3 Abstract ... 5 Samenvatting ... 6 Chapter 1. Introduction ... 7 Chapter 2. Methods ... 9 2.1 Literature review ... 9 2.2 Interviews ... 11 Chapter 3. Results ... 13 3.1 Literature review ... 13 3.2 Interviews ... 19 Chapter 4. Discussion ... 29 Chapter 5. Conclusion ... 34

Chapter 6. Research protocol ... 35

Chapter 7. Literature list ... 41

Appendix A - Interview questions ... 44

Appendix B - User satisfaction survey ... 48

Appendix C - Code tree interview results medical specialists and residents ... 51

(5)

Abstract

Introduction: Communication in the health care sector is an important aspect to continuity and safety of care for patients. For decades, medical specialists and general practitioners complain about deficiencies in their communication. In this study we examine how health information technology could improve the communication between the medical specialists/residents and general

practitioners at the university hospital Amsterdam UMC, location AMC.

Methods: We conducted a literature review and we performed interviews with medical specialists, residents and general practitioners. Based on the results of the literature review, the interviews and the possibilities within the hospital’s information system, we determined what health information technology could be implemented to improve the communication between the hospital and the GPs. We then proposed in a research protocol the design, implementation and evaluation of that health information technology.

Results: A total of fourteen medical specialists/residents and eight general practitioners were interviewed. The interviews show that ten out of fourteen medical specialists/residents agreed upon having an e-mail reminder that reminds them of sending letters. Results of the literature review mainly showed that a (discharge) template letter can be effective, but also that an e-mail reminder system and a shared health record can improve the communication. Given these results, an e-mail reminder fits best to our case.

Discussion/Conclusion: We present a plan for the design, implementation and evaluation of an e-mail reminder function within the hospital information system. Further research is needed to investigate what health information technologies could improve the communication between primary and secondary care to cover all aspects of deficiencies in their communication at hospitals with a hospital information system.

Keywords: improving communication, health information technology, primary care, secondary care, reminders

(6)

Samenvatting

Inleiding: Communicatie in de gezondheidszorg is een belangrijk aspect voor de continuïteit en veiligheid van patiëntenzorg. Al tientallen jaren klagen medisch specialisten en huisartsen over tekortkomingen in hun communicatie. In deze studie onderzoeken wij hoe zorginformatietechnologie de communicatie tussen de medisch specialisten, artsen in opleiding tot specialist en huisartsen in een academisch ziekenhuis Amsterdam UMC, locatie AMC, zou kunnen verbeteren.

Methoden: Wij hebben een literatuuronderzoek uitgevoerd en wij hebben interviews afgenomen met medisch specialisten, artsen in opleiding tot specialist en huisartsen. Op basis van de resultaten van de literatuurstudie, de interviews en de mogelijkheden binnen het ziekenhuis

informatiesysteem, hebben wij bepaald welke gezondheidsinformatietechnologie kan worden geïmplementeerd om de communicatie tussen het ziekenhuis en de huisartsen te verbeteren. Vervolgens hebben wij het ontwerp, de implementatie en evaluatie van die

gezondheidsinformatietechnologie beschreven in een onderzoeksprotocol.

Resultaten: In totaal werden veertien medisch specialisten en artsen in opleiding geïnterviewd. Daarnaast zijn er acht huisartsen geïnterviewd. Uit de interviews blijkt dat tien van de veertien medisch specialisten en artsen in opleiding instemden met het ontvangen van een herinnering per e-mail om brieven te versturen naar de huisarts. Resultaten uit het literatuuronderzoek laten zien dat een (ontslag)brief template, een e-mail herinneringssysteem en een gedeeld

gezondheidsinformatiesysteem de communicatie kunnen verbeteren. Gezien deze resultaten past een e-mailherinnering het beste in ons context

Discussie/Conclusie: Wij presenteren een plan voor het ontwerpen, implementeren en evalueren van een e-mailherinneringsfunctie binnen het ziekenhuisinformatiesysteem. Verder onderzoek is nodig om te onderzoeken welke gezondheidsinformatietechnologieën de communicatie tussen primaire en secundaire zorg zouden kunnen verbeteren om alle aspecten van tekortkomingen in de communicatie te dekken.

Trefwoorden: verbetering communicatie, zorg informatie technologie, eerstelijnszorg, tweedelijns zorg, herinneringen

(7)

Chapter 1. Introduction

1.1 Background

1.1.1 Improvement of communication in research

Communication in the health care sector is an important aspect to continuity and safety of care for patients [1]. For decades, medical specialists and general practitioners (GPs) complain about the deficiencies in their mutual communication [2, 3, 4]. For example, Kripalani et al, Bolton et al and Gandhi et al identified deficiencies in the communication between the primary and secondary care. These deficiencies varied from timeliness of letters to the quality of discharge letters. The

deficiencies in communication also impacts the continuity of care and the economy of health care organizations [4, 5].

Throughout the years attempts are made to improve the communication and timeliness such as implementation of protocols/guidelines, computer generated discharge template letters and the use of e-mail to send the discharge summary digitally [1, 4, 5, 6].

Nowadays there are still complains about the communication between the hospital and GPs [8, 9]. For example, according to a research from Mediquest in the Netherlands amongst GPs, GPs are not satisfied with the communication and information provision from the hospitals [10]. The GPs complain that medical specialists are not easy accessible and that letters from the hospitals are overdue or not sent at all. Another research performed in the Netherlands by van der Linden et al confirmed this by concluding in their research that the communication between GPs and the hospital is deficient [11].

1.1.2 The organization of health care in the Netherlands

In the Netherlands, GPs and medical specialists communicate via telephone contact, electronic and paper letters, joined post graduate courses and more recently via other communication platforms such as safe messaging apps and an electronic referral system (Zorgdomein) [8,12-16]. In the Netherlands, the primary means of communication between the hospital and GP is by letter. These letters inform the GPs about the discharge of the patient, change of medication, (ab)normal results or new policies. The Dutch committee of general practitioners (the NHG), stated in the guideline for information exchange between GPs and medical specialists that “a discharge letter should be sent in a timely manner to those who will take over the care of the patient to facilitate the continuity of care” [1]. The letters to the GPs are written by the residents and first need to be approved by their supervisors before the letters are sent to the GPs.

There are four groups of patients that visit the hospital. First, there are chronically ill patients who have to visit the outpatient department from the hospital so their health condition can be managed and controlled. Second, there are patients who are referred to the hospital for a short-term analysis and afterwards the consultation is closed. Third, patients who are on first aid department at the request of the GP. The patients stay there for a short-term. Fourth, there are patients who are admitted to the hospital and who are discharged after admission. During each of these events communication is required between the GPs and the hospital.

(8)

1.2.3. The Amsterdam UMC, Location AMC

The Amsterdam University Medical Center is an academic medical center which has two locations, AMC and VUmc. On 7th of June 2018, the AMC and VUmc merged and became one tertiary care

hospital [17]. This study took place in the AMC. Before the merger, the AMC tried to improve the communication from hospital to the GPs and implemented an e-mail based intervention called SnelleCor [18]. SnelleCor was a very successful intervention, but was discontinued when the Epic™ system (Epic Systems Corporation), a hospital-wide electronic health record (EHR), was

implemented. Medical specialists have asked for SnelleCor to be reimplemented in the EHR, but with the EHR there are many options for how we could implement it as well as other ways that we could address the broader problem of communication between the hospital and the GPs.

The aim of this study is to find out how communication between hospitals and the GP can be improved by health information technology. We will look at the AMC as a case. We will look in the literature to identify interventions that have been implemented to improve the communication between the health care professional and the GP. Next, we will use interviews to look at what medical specialists and residents perceive as barriers to effective communication with the GP and what solutions are preferred and vice versa. We will also propose a design for a quality improvement intervention to address those barriers and improve performance, and a study protocol for testing its effectiveness.

1.2 Research objective

The main question of this research is : how can communication between hospitals and the general practitioner be improved by health information technology?

In order to answer our main question, we developed the following sub questions:

 What interventions have been implemented to improve communication between the health care professional and the general practitioner?

 What do specialists see as the barriers to effective communication with the general practitioner and what solutions would they like?

 How would we design a quality improvement intervention to address those barriers and improve performance, and how will we test its effectiveness?

1.3 The structure of this thesis

This thesis consists of seven chapters. Chapter one is an introduction to this research project. Chapter two describes the methods used to answer the research questions. Chapter three contains the results of the literature review and the interviews. Chapter four and five contain the discussion and conclusion. We described our research proposal in chapter six. Chapter seven contains the references followed by the appendices.

(9)

Chapter 2. Methods

2.1 Literature review

2.1.1 Search strategy

We conducted a literature search using Medline and Embase databases. Articles were searched based on terms in the keywords, abstract, title and subject headings. We used a combination of synonyms and subject headings in five categories to design our search. Figure 1 shows the applied search strategy. The categories are general practitioner, communication, hospitals, electronic and intervention. We used the search terms to find interventional studies that test the effect of an intervention with the goal of improving the communication between the medical specialist, residents and the GP.

First the title and the abstract were screened by the primary researcher (UW). If the title and abstract matched the inclusion criteria, the article was read by the primary researcher.

2.1.2 Inclusion and exclusion criteria

The included articles were articles written in English that describe an attempt to improve the communication from hospital to the GP by using health information technology. The exclusion criteria were any studies that describe the move from paper-based to electronic communication as the only change.

2.1.3 Data extraction

The following data items were extracted by the primary researcher into a data extraction form: first author, year of publication, location where the study is performed, design, question/objective of the study, the identified problems in communication between GPs and medical specialists, methods, intervention, outcomes and whether the intervention was effective according to Roshanov. Roshanov defined an effective systems as “systems that improved primary (or 50% of secondary) reported outcomes of process of care or patient health” [19].

(10)

GP

1. (gp$ or primary care or family doctor$ or family physician$ or referring provider$ or general practi$ or family practi$ or outpatient physician$ or outpatient doctor$).ti,ab,kw,kf,tw. or general practitioner.sh. or general practice.sh.

COMMUNICATION

2. (correspondence as topic or interpersonal communication).sh HOSPITAL

3. (inpatient* or hospital* or ward* or ICU* or department of veterans affairs or intensive care unit* or specialist* or hospitalist* or acute care setting or specialty or secondary care or tertiary care or hospital* or medical center* or ambulatory care).ti,ab,kw,kf. or exp university hospital/ or exp hospital/ or exp hospital medicine/ or exp hospital department/ or exp hospital patient/ or exp hospital care/

ELECTRONIC

4. exp hospital information system/ or Electronic Health Record.sh. or Hospital Information Systems.sh. or exp Hospital Information Systems/ or exp Medical Records Systems, Computerized/ or exp computer systems/ or exp computer/ or exp information technology/

INTERVENTION

5. Clinical Trial/ or ((before adj5 after) or (pre$ adj5 post$) or improv$ or $intervention$ or control$ or trial$ or random$ or experiment$ or time series or nonequivalent group$ or effectiveness or ((follow$ or adhere$) adj5 (recommendation$ or advice))).ti,ab,kw,kf.

COMMUNICATION and ELECTRONIC

6. ((Health technolog$ or information technolog$ or computer$ or electronic) adj15 (communication or hospital-physician relation$ or letter$ or correspond$)).ti,ab,kw,kf. or (email$ or e-mail$).ti,ab,kw,kf,tw. or exp e-mail/ GP and COMMUNICATION and HOSPITAL

7. (discharge report$ or discharge summar$ or discharge letter$ or transition of care or hand off).ti,ab,kw,kf. ELECTRONIC and GP and COMMUNICATION and HOSPITAL

8. ((Health technolog$ or information technolog$ or computer$ or electronic) adj15 (discharge report$ or discharge summar$ or discharge letter$ or transition of care or hand off)).ti,ab,kw,kf.

9. (1 and 2 and 3 and 4 and 5) or (1 and 3 and 5 and 6) or (4 and 5 and 7) or (5 and 8)

Figure 1 - Search terms and search categories Items 1-5 in figure 1 represent the five categories of our search individually, and items 6-8 contain terms that imply more than one category. For example, item 6 contains terms that imply both “electronic” and “communication” (e.g. email).

(11)

2.2 Interviews

To supplement the findings from the literature, we conducted interviews with medical specialists and residents who are responsible for writing letters. The interviews were also conducted with the GPs who read the letters. The main goals of the interviews were to get the requirements for a system to improve communication with the GP and to learn what medical specialists and residents see as the barriers to good communication with the GPs (and vice versa), as well as what they see as facilitators and solutions to the barriers they mention.

2.2.1 Conducting interviews

For medical specialists, we used a convenience sample of contacts of the supervisors (SM and AV-K). For GPs, we did a web search on GP practices in the nearest region of the AMC. We then e-mailed the medical specialists and residents. The GPs were called and e-mailed. The GPs within the region of AMC were contacted, because they are more likely to refer their patients to the AMC and therefore need to communicate with AMC specialists. When a medical specialist or a GP (who was contacted via e-mail) did not respond in two weeks, he/she was reminded by a second e-mail.

The first researcher carried out the recruitment. We scheduled the interviews individually to best accommodate participants’ availability.

Semi-structured interviews were conducted by the primary researcher. Interviews with the medical specialists and residents were all performed in Amsterdam UMC, location AMC. The interviews with the GPs were performed at the practice of the participating GPs. Before the start of the interview, all interviewees were asked if the interview could be audiotaped and then transcribed. The number of interviews was determined by reaching saturation on the key questions, meaning having no new codes in the data [20]. The key questions are highlighted in appendices C and D.

2.2.2 Coding

The coding was done by the primary researcher. The interviewer used open coding and axial coding [21]. First open coding was done where all of the gathered data are read carefully and are divided into fragments. The pieces of text were then marked and labeled to cover the meaning of each fragment. Also if necessary, fragments were shifted to other codes that fitted best to the fragments. We then performed axial coding where the related codes were brought together under one code (main code), so the main code encompasses multiple sub-codes.

If during the transcription or coding process it was noted that an answer was incomplete (e.g. the subject was distracted and did not answer the question that was asked) then the subject was contacted afterward by email to clarify the answer.

2.2.3 Member checking

After analysis of all transcripts, an e-mail with a short synopsis of their answers on the questions about the system and letters was sent for feedback to the interviewees for member checking. We also sent a brief synopsis of the preliminary results of the interview. The medical specialists and residents also received a short summary preliminary results of the interviews from their own colleagues (medical specialists and residents). The GPs also received a short summary preliminary results of the interviews from their own colleagues (GPs).

(12)

2.2.4 Presentation of results

The results of the interviews are presented as a combination of recursive summarization (with codes in bold) and a presentation of the main codes that emerged from the interviews. The results are given per interview question. Comments that were relevant to the general topic of communication but were not direct answers to the questions were placed at the section called “other findings”.

2.3 Study protocol for proposed health technology intervention

The interviews and literature results were combined into a proposal for an intervention.This proposal was checked with a member of the hospital’s EHR for feasibility of implementation. A protocol for an interventional study testing the effect of this intervention is included as chapter 4 of this thesis.

(13)

Chapter 3. Results

This chapter presents the results of our literature review and the results of our interviews with the medical specialists, residents and GPs.

3.1 Literature review

3.1.1 Search and inclusion

The search resulted in 2413 articles, of which 6 were included. The details of the screening and inclusion are given in figure 2. Bibliographic screening resulted in no additional articles for inclusion.

(14)

3.1.2 Data extraction and synthesis

A summary of the included studies can be found in table 1. From these studies, five studies were performed in academic hospitals (Bischoff, Watkins, Dean, Medlock and Sheu) and one study in a secondary care hospital (Lampen-Schmith). Four studies (Dean, Bischoff, Sheu and Medlock) evaluated their intervention with a user satisfaction/experience survey. From these four studies there were two studies (Dean and Bischoff) that sent a survey to the sender of the template letters and the recipients of the letters (the GPs). Medlock only sent a survey to the medical specialists and residents and Sheu only sent a survey to the GPs. Furthermore, two studies (Medlock and Bisschop) measured percentage of letters completed on time. All other outcomes were measured by only one study (see table 1).

The studies all made an attempt to improve the communication from hospital to the primary care. The focus of these studies was improvement of the completeness and/or timeliness of the discharge letter. Four studies (Lampen-Smith, Dean, Watkins and Bischoff) presented as intervention a

discharge letter template. These templates are used to standardize the written communication with the GP. From these studies, there was one study (Bischoff) which had multiple interventions of which the standardized discharge template letter was part of the interventions. Lampen-Schmith’s study was the only study that did not show that a discharge template letter was effective. According to Lampen-Schmith it might be the result of the placement of the (proton pump inhibitor) prescribing information in the letter. The prescribing information was not seen by the majority of the GPs from the intervention group.

Watkins and Lampen-Smith both made an attempt to improve the health of the patients by improving the communication from the hospital to the GPs. These studies both tried to draw the attention of the GPs so the GPs would adhere to the advice from the hospital.

Two from the six studies (Medlock et Sheu) used other types of interventions than the discharge letter template. As intervention Sheu implemented a shared electronic medical record and Medlock implemented an e-mail based intervention that sends reminders to the medical specialists and residents to write a letter. Both studies proved that their intervention was effective. The

interventions of both studies sent notifications to the users. Medlock sent reminders and Sheu sent notifications to the GPs about the admission and discharge of patients.

Sheu’s study was the successor of Bischoff’s study. Sheu attempted to further improve the patient’s transition of secondary care to the primary care sector. Sheu concludes that the majority of the GPs still desired to have an additional direct contact with the medical specialists to ensure safe transition to primary care for the most discharges.

(15)

publication problems in communication between the hospital and GP Bischoff [42] 2011 United States A pre-post quasi-experimental study To increase the percentage of discharge summaries being completed on the day of discharge to at least 75% -Incomplete discharge summaries -Discharge summaries not completed within the day of discharge The timeliness and quality of discharge summaries were measured before and after the intervention using statistical analysis -Template based discharge summaries -Educational curriculum -Regular data feedback -Financial incentives -The percentage of completed discharge summaries rose from 38% to 83% on the day of discharge -The percentage of summaries that included all recommended elements increased from 5% to 88% The study did not prove if the health technology intervention did have a significant influence on the outcome. However, the interventions as a whole were effective. Lampen-Smith[43] 2012 New Zealand Randomized controlled trial To increase adherence with the New Zealand guidelines in -Inappropriate prescription of PPIs The use of 2-sided Pearson Chi-squared test A PPI-template was incorporated in the electronic discharge letter

The intervention did not show a signifcant difference between

(16)

patients discharged from hospital on proton pump inhibitors to compare the presence of a documented review adhering to the guidelines with required fields. intervention and control group Watkins[44] 2014 United States A pre-post quasi-experimental study To evaluate the effectiveness of the Emergency Provider Written Plan of Discharge (eEPWPD) template -Handoff communication deficiencies between the emergency department and GPs concering patients with low/risk chest pain -Patient records were investigated to determine whether patients were referred for further testing and whether follow-up with the PCP occurred after testing. -Use of statistical analysis Implementation of the Emergency Provider Written Plan of Discharge (eEPWPD) template

The follow-up rates in the intervention group increased significantly Yes Sheu[45] 2015 United States A pre-post quasi-experimental study -To understand cur-rent discharge communication practices and -Incomplete discharge summaries GPs and clinicians were surveyed

Shared EMR 52% of GPs were satisfied or very satisfied with

(17)

satisfaction of GPs towards the shared EMR. -To identify the areas in which communication can be improved -Discharge summaries were not received in time by the GPs after a shared EMR was implemented communication at patient discharge

GP = general practitioner, EMR = Electronic Medical Record Dean[46] 2016 United States of America A pre-post quasi-experimental study To design a discharge summary template and improve the quality of the discharge summaries Incomplete discharge summmaries -users of the template and primary care providers were surveyed Standardized discharge template 88% of the resondents (GPs) rated the new discharge summary (template) as better or much better Yes Medlock[18] 2017 The Netherlands Randomized controlled trial To improve the number and timeliness of letters sent from the hospital outpatient department to the GP using an email-based intervention evaluated in a Letters from medical specialists to GPs were not often written on time -Semi structured interviews -Statistical analysis -Surveys A monthly e-mail reminder

-The intervention group had 21% fewer patients without letters at 90 days after their visit

-The intervention group also sent their letters a median of 48 days sooner than in the control group

(18)

randomized controlled trial

(19)

3.2 Interviews

3.2.1 Formulating interview questions

Based on the goals of the interviews, interview questions were formulated by the primary

researcher, a researcher in Medical informatics (SM) and a clinician who is also a medical researcher (AV-K). This draft list of questions was combined with interview questions from similar, previously-published studies, and were reviewed again by the research team. The team revised the questions and agreed upon the final draft version. The first GP and the first medical specialist interviewed were also asked to provide feedback on the interview questions, to ensure that the questions were clear and covered all important aspects of the subject. Appendix A contains the interview questions. An example of an e-mail from SnelleCor was shown to the medical specialists and residents to inform or remind them of what the reminders looked like. For the GPs, an example of a letter from the

department of General Internal Medicine was shown. This example letter contains a core letter which is a brief overview of the most important content of the letter. The core letter is then followed by an extensive report. The core letter is a standard from the NHG.

3.2.2 Response rate and presentation of the results

Email invitations were sent to 21 specialists and residents, including an invitation to forward the invitation to colleagues. In total 9 specialists responded and were interviewed. An attempt was made to contact all general practices identified as being in the geographic vicinity of the AMC (n=28). Eight GPs responded and were interviewed. One GP invited a doctor in training to the interview, who also answered questions together with the GP. Another GP used to work in the region Amsterdam South East but did not at the time of the interview. The divergent results concerning the answer of that GP are marked with a *. The responses from this GP did not differ substantially from the other GPs, and thus are simply included with the data from the other GPs. All interviews were recorded except for one, because the recording device failed to record the interview. Immediately after that interview was performed, the answers to the interview questions were written down.

Saturation

Saturation was reached for the medical specialists and residents. For the interviews with the GPs, saturation was not reached but no additional GPs responded to the invitation for interviews. Member checking

From the hospital, twelve out of fourteen interviewees responded and gave their feedback on the summary. From the general practice, three GPs responded.

Below, the summaries of the answers are given per interview question. The questions that were both asked to the GPs and medical specialists and residents were put together in section 3.2.3. The results of the interviews are presented as a combination of recursive summarization (with codes in bold) and a presentation of the main codes that emerged from the interviews.

(20)

3.2.3 Interview results medical specialists, residents and GPs

The ways medical specialists communicate with general practitioners and vice versa

All medical specialists mentioned that they communicate via letters and telephone calls with the GPs. From the fourteen medical specialists, four mentioned having e-mail contact with the GPs alongside their regular information systems.

The GPs mentioned that they have contact through their referral system and also via telephone. Next to that, seven GPs mentioned that they sent e-mails to medical specialists which is aside from the regular communication channels. Also the GPs receive letters from the hospital via their information system and via (digital) post.

The ideal communication between the medical specialist and the general practitioner and vice versa The most common ideal situation mentioned by the medical specialists and residents (six

interviewees) is a shared health record where the GPs have access to the hospital information system. Appendix C contains the code tree with more examples of codes of ideal situations. A number of possible ideal situations were mentioned by the GPs. Three GPs mentioned the use of e-mail as an ideal medium for contact. Appendix D contains the code tree with more examples of codes of ideal situations.

3.2.4 Interview results medical specialists and residents

Reasons that letters are sent with delay or not sent

A total of fifteen different reasons of why letters are sent with delay were mentioned (table 2). There were two categories of reasons mentioned most by the medical specialists: lack of time and logistics. Considering the category logistics, medical specialists and residents explained that the letters first have to be sent to the supervisor for approval before the letters can be sent to the GP. Another medical specialist explained that discharge letters from patients who are discharged on Friday nights and in the weekends are sent after the weekend.

Two interviewees mentioned that a reason for letters sent with delay is that writing the letter is not a priority: “Well it is a policy that you write a letter with every new patient and once a year for the chronically ill patients. Sometimes, definitely on an assistant clinic, you run an outpatient clinic for four months and then give it to someone else. If you see a person once ‘between nose and lips’ and it was very busy, then you do not feel that you should immediately write a letter about that visit. If your successor does the same four months later and his successor also, then one year is already over.”

One interviewee mentioned that the reason letters are sent with delay is because of lack of discipline: “I cannot say that it is only a lack of time because we do patients with the same timeframe. So I think it's a bit of discipline.”

Also administrative mistakes were mentioned. For example, one interviewee explained this and said: “the patient might have left from one GP to another GP and the address of the former GP is still in the file with addresses.”

(21)

Table 2 Reasons letters are sent late

Reasons letters are sent late Number of medical specialists and residents who mentioned this

Lack of time 7x

(due to the) logistics 6x

Relevant information cannot be given within the set term

3x

Supervisors are late with approval of the letters 2x

The administrative burden 2x

Writing a letter is not a priority 2x

lack of discipline 1x

The hospital information system does not give the right support with regards to writing a letter fast

1x

General practices are not digitally connected 1x Administrative mistakes with regards to the

address of the GP

1x

New medical specialists in trainee need to get used with working with the hospital

information system

1x

Rapidly changing residents 1x

The workload 1x

Medical specialists forget to write letters 1x There is no need felt in writing a letter because patients often call the hospital for explanations

1x

Patients call the medical specialist for questions so there is no need felt in communication with the GP

(22)

How letters sent with delay can be prevented

There were thirteen different categories of ideas mentioned about how letters with delay can be prevented (table 3). There was only one medical specialist who mentioned that there is no solution to help send letters on time because of the current policy that doctors in training first have to send their letter to their supervisor before it can be sent to the GP.

Two categories of ideas were mentioned most (four interviewees): the reduction of the administrative burden and that the letters should be modified. About the reduction of the administrative burden the interviewees mentioned that the secretary should assist more by doing extra tasks. For example one interviewee mentioned “a better support from the secretary so that there will be a better secretarial support and that there will be people who can measure the blood pressure and can organize logistic things”. Another interviewee said it is better to “start with non medical issues being taken care of by someone else (other than the medical specialist)”. Concerning the modification of the letter, the medical specialists and residents mentioned that there should be less focus on the details of the letter and that the letters should be shortened. Another interviewee mentioned that the hospital should act just with regards to sending the letters when it is necessary saying: “To give an example, the guideline says that you have to send a letter every year and the AMC has adopted it as a policy. But that is a bit of a stupid rule. Some people have just 5 years diabetes and 5 years a controlled Hba1c, so the following year you send [to the GP] ‘it was good again’. That is a bit silly to do so. ”

Another interviewee mentioned that awareness (create awareness) among the supervisors must be raised so they know they have to write a letter.

Table 3 How letters sent with delay can be prevented

How letters sent with delay can be prevented Number of medical specialists and residents who mentioned this

The letter should be modified 4x

Reduction of the administrative burden 4x Improve the usability of the hospital

information system

2x

Writing letters should have a higher priority 2x

Place reminders 2x

The hospital should act just with regards to sending the letters when it is necessary

1x

Create awareness 1x

Residents who are in their last year or second last year do not have to send their letters to the supervisor for approval

(23)

The hospital should be more strict 1x The use of a shared record for GPs 1x The availability of an overview of letters that

need to be written

1x

Hire extra colleagues 1x

Patients should check and modify (if needed) their personal data by each visit to the hospital

1x

How the system should operate

There were ten out of fourteen interviewees who did support the re-implementation of a system like SnelleCor. There were different answers on how the system should operate (table 4). The four other interviewees who did not support this function mentioned the following: “it reminds the user that he/she is running behind, there is already a list provided with letters that need to be sent with the performance of other colleagues, e-mails should not be sent because people are getting crazy about e-mails”. One interviewee explained that there already is such a function in the hospital’s EHR that shares an overview of letters that need to be sent.

Receiving the reminders via Inbasket was mentioned the most (by seven interviewees). Inbasket is the mailbox within the EHR, where the medical specialists and residents communicate with each other and other health care professionals from the hospital (for example a laboratory technician) about patients using patient data. There were two interviewees who had no preferences in via which channel the reminders should appear.

Three interviewees mentioned that it should be possible to turn off some reminders for specific groups of patients or just for one patient. It was also mentioned by one medical specialist that there should be a check whether the letters have been sent after the reminders are generated and another interviewee mentioned that the reminders should be sent to the right medical specialist.

Table 4 The way via which the reminders should be sent - medical specialists versus residents

Number of medical specialists who preferred this (N=5)

Number of residents who preferred this (N= 5)

Via Inbasket 4 3

No preferences 0 2

(24)

The frequency of the reminders

The answers on how often a reminder needed to be generated varied highly among the interviewees (Table 5). One medical specialist did not answer this question.

Table 5 The preferred frequency of reminders

Frequency Number of medical

specialists who agreed upon this (N= 5)

Number of residents (N=5)

Daily 0 1

Monthly or once in two months

1 2

Every two weeks 1 0

Once in 2-3 months 1 1

Once a week (inclusive second reminder two weeks later)

1 0

Once a week 0 1

No answer 1 0

Term within discharge letters and letters from the outpatient clinic have to be sent

Also the answers on this question variates highly among the interviewees (6a-c). One medical specialist preferred to send a letter once in two years instead of once a year for some chronically ill patients (CP). Also one medical specialist mentioned for the chronically ill patients (CP) that once a year a letter needs to be sent or when important information needs to be sent to the GP. The answers from some interviewees were unclear, because there was no distinction made between patients from the outpatient clinic and from inpatient clinic. So these answers to the questions were stated as “not clear”.

Table 6a Term within outpatient letters need to be sent to the general practitioner for chronically ill patients (CP)

Answer Number of

times mentioned

Once per year 6

Not clear 8

(25)

Table 6b Term within outpatient letters need to be sent to the general practitioner for new outpatients (NP)

Table 6c Term within discharge letters need to be sent to the general practitioner

When discharge letters need to be sent to the general practitioner

Number of medical specialists (N=9) Number of residents (N = 5) Within 24 hours 4 3 As soon as possible 1 0 Within 24 or 48 hours 2 0 Within 48 hours 0 2

Within three days 1 0

Within two months 1 0

Other findings

There were two medical specialists who mentioned that medication was not accurate in the template generated by the hospital’s EHR. More codes of other findings are presented in the code tree of Appendix C.

Answer Number of times

mentioned

At the patient’s second visit 1

Six weeks 1

When diagnoses can be made and/or when the treatment plan is made

1

Within two weeks 1

(26)

3.2.5 Interview results general practitioners

Receiving letters with delay or not receiving a letter

All GPs agreed that they receive overdue letters and two medical specialists even mentioned that they do not always receive letters from the hospital.

The effect of receiving letters with delay on patient care

There were six GPs who mentioned that the delay of letters results in them having a lack of

information. For example, two GPs explained that they have to ask their patients about their visit at the hospital and about what was discussed during that visit. Appendix D contains the code tree with more examples of codes of effects of receiving letters with delay.

The term within the GPs want to receive a discharge letter after the patient gets hospitalized

The answers on this question varied from “within 24 hours” to it “depends on the context” (table 7).

Table 7 The term within the GPs want to receive a discharge letter after the patient gets hospitalized

Number of GPs specialists who chose this frequency (N=7)

Within 24 hours 2

It depends on the context 3

Within two weeks 1

If the GPs is expected to do something, than as soon as possible. But otherwise within two weeks.

1

How often the GPs want to receive letters about chronically ill patients

Table 8 The term within GPs want to receive letters about chronically ill patients

Number of GPs specialists who chose this frequency (N=7) Directly after a change in the treatment

policy

3

Once a year* 2

For patients who visit the specialist once a year, one letter per year.

2

When the patient’s health is stable, then once a year or once per two years

(27)

It depends on the disease the patient is diagnosed with

1

Once or two times per year 1

The timeframe within the general practitioner wants to receive a letter about a patient who has been seen on the outpatient clinic for the first time

In total, four GPs mentioned that the speed in which they want to receive the letter depends on the context of the case. There were two GPs who wanted to receive a letter right after the lab results are in. It was mentioned that for chronically ill patients who regularly visit the hospital, a letter can be sent once a year.

Number of GPs specialists who chose this timeframe (N=7) Depends on the context of the case 4

Right after the lab results are in 2 For patients who visit the specialist once a year, one letter per year.

1

It seems like here we’ve suddenly changed from the format of summaries with the code words in bold to code words as headings. I think the code words in bold works better, but in any case it should be consistent.

The content of the letter

Four GPs mentioned that the letters were too long. There were also two GPs who mentioned that the length of the letters differ per specialism.

The format of the letters

All GPs gave positive remarks about the format of the letter template. One GP mentioned that it is much safer, because it is difficult to miss something. “If you receive a lot of text, you read a little diagonally because everyone is under time pressure. And this is just the core”.

There were also GPs who gave advice for improvement. Two out of seven GPs mentioned that the format of the letters differ. Furthermore, another GP mentioned that not only the most recent letter gets sent, but also other letters from previous correspondence. This GP prefers to receive only the most recent letter. Another remark was that the format changes when the letters are sent digitally. So the letters sent from the hospital look different than the letters the GPs receive. Barriers in the communication between the medical specialist and the general practitioner Lack of time was mentioned by three GPs as a barrier to communication. Two GPs explained that they experience that the doctor on duty has not much time to discuss a patient when the GP does not call concerning an emergency. Also one GP noticed that the medical specialists are under time pressure: “Sometimes I think he (the specialist) has to see five more patients at the outpatient clinic.

(28)

The specialist does not have time for me”. Appendix D contains the code tree with more examples of codes about the perceived barriers in the communication.

Facilitators to communication with the medical specialist

Several facilitators of communication with the medical specialists and residents were given. Knowing the medical specialists (personal contact) was mentioned by all GPs as a way to facilitate the

communication. Five GPs mentioned post-graduate courses together with the medical specialists as facilitator to the personal contact. Other examples of codes for facilitators to communication can be found in Appendix D.

What GPs see as important when it comes to their relationship with the medical specialists

The following aspects were mentioned: personal contact to get to know each other (4x), collegiality (3x), ease of access (2x), timely feedback (1x) and the competence and knowledge from the medical specialists (1x). For example, one GP mentioned that medical specialists should understand what the GP needs and another GP mentioned respect towards each other.

How GPs describe their experience with the secondary care

In total, five GPs mentioned that smaller hospitals have a better ease of access to medical

specialists than big hospitals. Also one GP mentioned that letters are received faster when they are sent digitally. There are also differences in the communication per hospital according to one GP. It was even mentioned by two GPs that there are differences in the length of the letters per

specialism. Furthermore, one GP mentioned that the AMC is less accessible compared to other hospitals. It was also mentioned once that there are differences in the length of letters per hospital.

Other findings

One GP mentioned that it takes too much time to process a long letter. Also another GP mentioned that the letters of mental health care organizations contain irrelevant information and do not contain a desirable format according to one GP. More codes of other findings are presented in the code tree of Appendix D.

(29)

Chapter 4. Discussion

4.1 Main results

In this study we aimed to investigate ways to improve the communication between the medical specialists/residents and GPs by the use of health information technology. In the systematic review (chapter 3), a total of six studies were included that implemented and tested a health information technology intervention on its effectiveness. The interventions that we found were four discharge summary templates, an e-mail reminder system and a shared health record. The results from the literature were supplemented with interviews held with medical specialists, residents and GPs, which showed that there are still deficiencies in their communication where the timeliness of letters was mentioned as one of the deficiencies (chapter 3). By conducting a literature review and

performing interviews with fourteen medical specialists/residents and eight GPs, we determined what health information technology could be implemented to reach our aim. Ten out of fourteen medical specialists/residents agreed upon having an e-mail reminder system that reminds them of sending letters.

4.2 Strengths & Limitations

This research has several strengths. First, we performed a systematic literature review that presents an overview of health information technologies that were implemented to improve the

communication between primary and secondary care. Second, we searched in two databases (Medline and Embase) that cover a great portion of published articles. Third, terms that we pulled from articles, their subject headings and their synonyms were used in our literature search. This reduced the chance of missing relevant published articles. Fourth, we did not use our own definition, but a definition from literature (Roshanov et al) to determine when an intervention was successful. Fifth, we did not randomly collect data from the included articles but we used a structured data extraction form to extract the data. Sixth, this study also presents the current perceptions of the communication between the medical specialists, residents and the GPs within the region of the AMC containing their perceived barriers and facilitators to their communication. Seventh, we performed member checking to improve the accuracy of the interview results and to check the validity of the interpretation of the interview questions. Eighth, we used open questions during our interviews to provide an opportunity for the interviewees to share more relevant information. Ninth, we tried to maximize the response rate by performing the interviews at the offices and practices of the medical specialists, residents and GPs. Tenth, the interview questions were reviewed by three researchers, one GP and one resident to make sure the questions were clear to the interviewees. Furthermore, we performed a pilot check on the interview questions and recorded the interviews to minimize the loss of relevant data. We aimed to reach saturation on the key questions of the interviews and did reach saturation with the medical specialists and residents. We performed a deviant case analysis paying attention by mentioning results from the interviews that did not fit the picture we were drawing.

Although several strengths are mentioned, there are also limitations. First, our search may have missed some articles, however our systematic approach should minimize this. Second, we excluded articles written in English, however the most important articles are written in English. Furthermore, the literature review was performed by one researcher. The interviews, coding and analysis of the interviews were also performed by one researcher. If these steps were done by two or more

(30)

researchers where for example, two researchers were coding the transcripts and analyzed the codes on their own and afterwards discussed upon that, the coding could have been more complete and of better quality. It might be that one researcher missed an important item. This item could have been identified by the other researcher. Data saturation was not achieved for the interviews with the GPs, the GPs did not respond or declined to participate. For the interviews, we chose to send a brief summary of the study results to our interviewees in our member checking step. This means that the member checking step could also have functioned as a consensus round, similar to that used in Delphi panels [22]. However, subjects were not invited to change their answers based on the preliminary study results and no participant indicated that they wanted to do so. We asked the medical specialists and residents what their feelings are towards SnelleCor, which is one intervention. This may have led to preferences for using Snellecor, however we asked the open questions before we asked about SnelleCor and the answers on the last two questions did not indicate that the questions are biased because of other interesting suggestions provided by the interviewees. We cannot mention much about the term within certain letters have to be sent to the GP because not all medical specialists and residents made an explicit distinction between recheck patients and new patients and if the subject did not note the distinction, the interviewer did not press for specific answers. Last, we used a convenience sample of medical specialists and residents which might have led to a response bias where the interviewees gave answers in favour of this current study.

4.3 Relation to other work

4.3.1 Literature review

To our knowledge, chapter 3 of this thesis is the only literature review that searched for health information technologies that improve communication between the hospital and the GP. Prior work in this area includes literature reviews that included studies that did not implement health

information technology interventions [4, 23]. For example, Kripalani et al presented in a systematic review an overview of interventions implemented to improve communication between secondary care and primary care [4]. These interventions were not all health information technology

interventions. Some interventions tackled the quality of discharge summary letters and others timeliness of the letters. Similar results with our interview results are dissatisfaction about the timeliness of the letters resulting that patients or their families have to inform the GP about the hospitalization, the need to get (discharge information) when medications have been changed. The studies found in our literature review were not included in Kripalani’s review because these were published after Kripalani’s review was done.

4.3.2 Interviews

Results from our interviews were confirmed by previous literature and our literature study. These results confirm that the communication problems experienced by doctors in our setting are similar to those in other settings. Our literature study showed that there were deficiencies in

communication between the hospital and the GPs. These were the timeliness of the letters sent to the GP and the incompleteness of (discharge) letters. These deficiencies were also mentioned as barriers at our interviews.

From our literature review, the two studies from Watkins and Lampen-Smith tried to improve the healthcare of the patients by making sure the advice to the GP is not forgotten in the letter and also by making the advice more clear by implementing a template. The results of our interviews also

(31)

show that there needs to be a clear header in the template letter with advice to the GP so the GP would not miss important information. This might also have solved the deficiencies in the

communication Lampen-Smith and Watkins faced.

Also the need for timeliness of communication is a known need for improvement in previous literature [24, 25]. Next, both GPs and medical specialists/residents talked about the need for improvement of the ease of access to one another. This aspect was also mentioned in other literature [8, 24, 26, 27]. For example, Berendsen et al investigated how the GPs and medical specialists from the Netherlands value their mutual communication through a survey. Their results show that less than half of the medical specialists were satisfied with the accessibility of the GP via telephone whereas three quarters of the GPs were satisfied about the telephone accessibility of the medical specialist.

Also the desire for a shared health record, which was mentioned by six medical specialists and residents and only one GP, has also been confirmed by other literature [25, 27] and also by Berendsen et al.

There was one GP who mentioned that residents from certain departments should visit a general practice to understand the role/workplace of the GP. This was confirmed by interviews with GPs in other literature such as Berendsen et al, however the GPs in Berendsen’s study were talking about the medical specialists [8, 26]. They did not specifically mention residents. What also was confirmed by three of Berendsen’s studies among GPs and medical specialists in the Netherlands, is the need for GPs to know the medical specialists and residents [28, 29]. This was also mentioned by medical specialists and GPs at the study of Sampson et al [26]. In our study, none of the medical specialists mentioned this factor as a facilitator to their communication or collaboration with the GPs. We might have a different result because we did not ask them about barriers and facilitators in their communication with the GP.

Next, one medical specialist was not satisfied about the content of referral letters. This was also confirmed by Berendsen et al where two-thirds of the medical specialists shared the opinion that the quality of the referral letters was insufficient. The number of medical specialists was low. This might also be due to the fact that there was no specific question asked about the quality of the referral letter.

There were also confirmations from literature concerning results from our interview with the GPs. Berendsen confirmed in her study that GP’s are satisfied with the specialist’s letter and that there is a need among GPs for post-graduate courses with medical specialists to facilitate the

communication [8]. Stille et al and Etesse et al performed a qualitative study sending surveys to GPs from France and New England and both confirmed that GPs prefer to receive a phone call about a patient’s admission and urgent communication [24, 30]. Stille et al also confirmed that the GPs often have a lack of information about their patient’s visit or stay at the hospital. Farquar interviewed GPs from England to get insight on their views about communication issues between the primary and secondary care in relation to ovarian cancer patients [25]. Farquar et al confirmed that GPs also preferred the use of e-mail to improve the speed of communication.

Jones et al investigated through interviews and focus group sessions the perceived challenges and solutions for improving coordination of care among GPs and hospitalists in the United States. As in

(32)

our study, Jones et al also confirmed that GPs prefer to be notified about medication changes [27]. Jones also confirmed that GPs would like to reach the correct specialist in cases when they call the hospital [27]. Last, Berendsen et al also confirmed that GPs prefer to experience mutual respect when it comes to the collegiality [29].

4.4 Interpretation and implications

Results of this study show that letters are not sent on time and this might be because it is not seen as a priority. It might not be seen as a priority because of the mentioned workload and lack of time. But it also might be because medical specialists and residents are waiting for lab and search results or there is no relevant information to tell.

The literature review shows that four of the six studies implemented an intervention that has already been implemented at our institution: a standardized discharge template letter.

The two other studies already had a discharge summary template letter and made an attempt to improve the communication by implementing another type of intervention. One of these two studies used a shared electronic health record where GPs have access to their patient’s hospital data. This is also mentioned most by the medical specialists and residents (total of six) as an ideal situation.

However, in the Netherlands, an attempt of implementing a shared medical record was rejected by the government in 2011 [31, 32]. This idea of having a shared medical record was not only rejected by the government, but also by the Dutch citizens [33]. This shared medical record was not meant to function as a national electronic record, but contained the functionality where all health care providers could ask and get patient information when needed for (further) treatment of the patient [31].

Knowing that this service in the current state of healthcare, needs time for acceptance of such an intervention, it was wiser to choose an intervention that contains a small step to such an ideal situation, which is the intervention Medlock presented in her study. Although this intervention only improves the timeliness of the letters, it still is a step towards improvement of an aspect of the communication. The results of the interviews showed that the majority (ten out of fourteen) of medical specialists and residents agreed to have a system like SnelleCor that reminds them of writing letters, we chose this intervention as best intervention for this context.

We performed a literature review that presents health information technologies that aimed to improve the communication between the GPs and the hospitals that have a hospital information system. This may help other researchers to do further research on this subject to see if the results of our study are confirmed in future studies. Second, this research confirms for hospitals who already have a template letter that implementing an e-mail reminder system could help improve the communication. Third, the results of the interviews present useful insights to the Dutch society of general practitioners (the NHG) and the Federation of Medical specialists (Federatie Medisch specialisten) who write guidelines for the information exchange between primary and secondary care. Fourth, the satisfaction of the GPs with the hospital’s template letter is presented together with suggestions for improvement which are also useful to organizations such as the NHG and the Federation Medical specialists. The results of the interviews also provide current perceptions of the communication between the AMC and the GPs within the region of the AMC containing their perceived barriers and facilitators to the communication. The results of the interviews were confirmed by other studies, so the results of this study can be generalizable to other countries.

(33)

4.5 Future research

There is limited research on the use of health information technology to improve the communication between GPs and hospitals. The next step should be to study the effectiveness of the e-mail

reminder to improve communication, as we present in our research protocol in chapter six. Although we did not assess the effectiveness of a reminder system on the timeliness of letters between primary and secondary care, we still recommend research about the effectiveness of e-mail reminders outside the Netherlands to assess the effectiveness of the intervention. Our proposed intervention does not cover all deficiencies in the communication between the primary and secondary care. So, we also recommend that more research is needed to have more insight in how health information technology could cover all aspects of communication.

(34)

Chapter 5. Conclusion

Based on the results of our literature review and based on the interviews performed, we propose that a reminder function within the hospital information system could lead to an improved communication between primary and secondary care. The majority of studies from the literature review looked at implementation of templates, and only two studies looked at improvements suitable for hospitals where templates are already in place. The results of the interviews show dissatisfaction about the communication between the primary and secondary care from which a number of communication deficiencies were mentioned in previous literature. The GPs and medical specialists/residents both were dissatisfied with the ease of access to each other.

As facilitator to the communication the medical specialists and residents preferred writing short letters and GPs preferred having post-graduate courses with the medical specialists and residents to get to know them. In the long run, a shared patient record is desired by both the specialists and GPs that were interviewed. This should be investigated further. Also further research is needed to investigate what health information technologies could improve the communication between primary and secondary care to cover all aspects of deficiencies in their communication at hospitals with a hospital information system. In the next chapter we present a plan for the design,

(35)

The results of Chapter 2 and 3 were discussed with two experts in the technical capabilities of the hospital’s EHR, to arrive at the intervention design that is used in the following research proposal.

Chapter 6. Research protocol

6.1 Introduction

As outlined in the general introduction of this thesis, communication between GPs and hospital specialists remains an area for improvement. The results of our interviews (chapter 3.2) also identified deficiencies in the communication between GPs and the hospital. GPs mentioned that they still get overdue letters. There were also GPs who mentioned that letters are sometimes not sent. The results of the interviews also show several consequences of receiving overdue letters on patient care. These consequences are that the GPs are having a lack of information, they are obstructed in writing the treatment policy, the GPs have to ask the patients about what was discussed during their consult or stay at the hospital, a wrong treatment policy might be written, certain tasks might be performed twice and the continuity of care gets obstructed.

The literature review (chapter 3.1) shows that there is little research on improving communication between the hospital and GPs by implementing health information technology in hospitals where the improvement did not consist of moving from paper-based to electronic communication. Another e-mail based intervention was SnelleCor. SnelleCor was a very successful e-mail based decision support intervention that sent reminders to medical specialists and residents about letters that still needed to be sent to the GPs [18]. Aspects of SnelleCor that the authors felt contributed to its success included that the medical specialists and residents use their hospital email account daily, so they could read and see the e-mails sent by SnelleCor. The medical specialists and residents were then able to use the reminders as a measure of their performance. The implementation of a

hospital-wide patient record system eliminated standalone systems like SnelleCor and medical specialists have asked for SnelleCor to be reimplemented. However, an integrated patient record also provides functionality that was not available in the previous architecture.

In our previous work, we gathered opinions of medical specialists and residents on communication problems and possible solutions1. These have been formulated into a proposed intervention. The

aim of this research is to test the effectiveness of this intervention on the percentage of overdue letters and the timeframe between the patient’s visit and the day the letter is sent to the GP. Our secondary aim is to evaluate the users’ perspective on this intervention. This intervention differs from SnelleCor mainly because the intervention is within the integrated patient record. First, in the time that SnelleCor was used, the hospital did not have a hospital-wide electronic patient record system. Our proposed intervention will be researched in a hospital that does have a hospital-wide electronic patient record system. Second, SnelleCor sent e-mails to the users and our proposed intervention is part of the patient record system. So to add knew knowledge to this field, we will implement, test and evaluate a health information technology system in a tertiary-care, university medical center that has an integrated patient record system.

We have the following hypotheses for our primary and secondary outcomes:

(36)

Primary outcome hypotheses

 There is a decrease in overdue letters after the start of the intervention.

 There is a decrease in the timeframe between the patient’s visit and the day the letter is sent.

Secondary outcome hypotheses

 The majority of the respondents from the user satisfaction survey will show a positive remark on this intervention.

6.2 Methods and study design

6.2.1 Study design

To achieve our goals we will perform a quasi-experimental interrupted time series design, which is the best quasi experimental design [34], and we will send a survey to all participants to evaluate the satisfaction of the users.

6.2.2 Setting

The study will be performed at a tertiary-care, university medical center (Amsterdam UMC). This university medical center has two locations where patient care is given. This study is only performed in location AMC. From this location, the outpatient departments from Division A (except for the Dermatology department) will have the intervention implemented. Division A stands for the internal and external specialisms.

6.2.3 Participants and data collection

All medical specialists and residents from division A will be included except for the department Dermatology. Routinely collected data from 12 months before and 12 months after the start of the intervention will be analyzed, as this is a minimum number of data points required Wagner et al. [35].

6.2.4 The intervention

Within the hospital’s EHR it is possible to send reminders to the medical specialists and residents via the EHR’s internal mail system (“InBasket”).

The intervention will contain a reminder mail for each overdue letter. These reminders will be sent to the InBasket and will be placed in a subfolder named overdue letters (figure 1).

The reminders will contain information about which letter from which patient needs to be written. The reminder also contains two links. One link is to open the consult that generated the reminder. And another link is to open a new encounter to write a letter which is not associated with a specific visit. The clinicians can also suppress the reminders for a selected number of days, although this requires clicking through several menus. The users can also click on the cancel/finished button that permanently removes the reminder.

Referenties

GERELATEERDE DOCUMENTEN

In cartesian coordinates and Euclidian space This tensor equation is valid for all coordinates Covariant derivatives. Take covariant

Ook door de Commissie van de Europese Gemeenschappen (2000) wordt gesteld dat alle strategieën voor leven lang leren in de eerste plaats gebaseerd moeten worden op de

applying the MASS intervention, students with increased medical absence rates (i.e. four periods in 12 school weeks, or more than six consecutive days) are identified by the school,

Al- though one third of the GPs mentioned good collabor- ation as a stimulating factor for referring more patients to movement facilities, regular meetings of health professionals

Methods: We analyzed the influence of eight practice characteristics, such as type of practice, percentage female general practitioners, and employment of a practice nurse, on

Practice nurses from intervention practices (1) visit older people at home for a comprehensive assessment of their health and well-being; (2) discuss results with the GP; (3)

One of the general insights gained from this study is that the answer to what Dutch consumers find most important regarding online consultations is not an obvious

As a result, both specialties were better able to find specific results (such as notes) of other specialties, thereby increasing the Mutual awareness between these