Reasons for not reporting patient safety incidents in general practice: A qualitative study

Abstract

Objective

To explore the reasons for not reporting patient safety incidents in general practice.

Design

Qualitative interviews with general practitioners and members of the project group.

Setting

General practice clinics in the Region of Northern Jutland in Denmark.

Subjects

Twelve general practitioners.

Main outcome measures

The experiences and reflections of the involved professionals with regard to system use and non-use.

Results

While most respondents were initially positive towards the idea of reporting and learning from patient safety incidents, they actually reported very few incidents. The major reasons for the low reporting rates are found to be a perceived lack of practical usefulness, issues of time and effort in a busy clinic with competing priorities, and considerations of appropriateness in relation to other professionals.

Conclusion

The results suggest that the visions of formal, comprehensive, and systematic reporting of (and learning from) patient safety incidents will be quite difficult to realize in general practice. Future studies should investigate how various ways of organizing incident reporting at the regional level influence local activities of reporting and learning in general practice.

Previous studies have reported positive attitudes to incident reporting among Danish GPs.

• The results suggest, however, that formal reporting of patient safety incidents may be more attractive to GPs as an idea than as an integrated activity in a busy clinic. While most GPs in this study were initially positive towards the idea of reporting, they actually reported very few incidents.

• The reasons for the low reporting rates were found to be a perceived lack of usefulness, issues of time and effort, and considerations of appropriateness.

• Future research should compare how various ways of implementing and organizing incident reporting influence local activities of reporting and learning.

Introduction

During the last decade issues of patient safety have come to occupy a prominent position on the political agenda in health care, and in several countries systems for incident reporting have been promoted as a means to improve patient safety [1]. Such systems have mostly been implemented in the hospital sector, but recently steps have been taken to include the primary sector as well. In Denmark, reporting patient safety incidents has been mandatory in the hospital sector since 2004 [2], and in 2009 this legislation was extended to the primary health care sector with effect from 2010. This development raises the question of how professionals in general practice (and other parts of the primary health care sector) will respond to efforts to institutionalize incident reporting. Previous research has found that general practitioners in Denmark express positive attitudes towards formal registration and learning from patient safety incidents [3,4]. But while concepts and technologies related to quality and patient safety are by definition loaded with positive values, which make them difficult to oppose at an abstract level, their actual translation into daily practice is rarely a smooth process as challenges often emerge when the various implications of adopting specific technologies appear more clearly to the various users involved [5–7]. And although research on technology acceptance generally purports a tight relationship between intentions and actual use [8] this relationship is not a deterministic one due to the time factor and early user experiences [9]. Therefore, in order to establish a realistic impression of the prospects, conditions, and challenges for implementing formal and systematic reporting of (and learning from) incidents in general practice, it is necessary to investigate specific attempts to promote such reporting. Against this background we conducted a qualitative case study of a regional project for incident reporting in general practice. While a small number of GPs initially expressed enthusiasm and signed up to use the system actual reporting activity turned out to be low in these clinics, and the aim of the paper is to explain the low reporting rates among these (presumably) “early adopters” [11].

Material and methods

The study was conducted in the Region of Northern Jutland where a regional project on incident reporting was in operation from 2008 to 2010 [11,12] (Table I). From June 2009 to October 2010 semi-structured interviews [13] were carried out with 12 GPs from nine different clinics (Table II). All interviews were performed in the clinics and all interviews were individual interviews (except from one group interview with four GPs from the same clinic).

Table I.

The incident reporting project.

Purpose	The overall purpose of the project was to prepare general practice for the (expected) national legislation on patient safety incident reporting (making incident reporting mandatory for all health professionals in primary care) by trying out and developing an organizational model for working with incident reporting in general practice
Project management	All GP clinics in the region (approx. 200) were invited (via e-mail and a regional GP journal) to report patient safety incidents observed in the clinic, in other primary health organizations, and in the secondary sector. Also, in order to promote the concept of incident reporting and demonstrate how to use the project website, the project manager participated in regional and local GP meetings, and visited individual clinics. Due to the primary purpose of the project, the emphasis of the project manager was not on achieving widespread regional adoption, but to increase awareness of incident reporting, and to help interested clinics get started with using the reporting system
Reporting, analysis, and feedback	The core element in the reporting system was the website, which contained a short reporting formula for supplying information on the particular incident (what happened, where, and what were the consequences) and suggestions for preventive actions. Also, the website hosted general information on the project, including statistics and newsletters summarizing the themes of the submitted reports. Upon receiving a report, the project manager categorized it according to severity and place of origin. If the report concerned an incident in general practice the project manager sent a receipt to the clinic acknowledging the report and possibly asking for permission to use the case in newsletters and for educational purposes. Also, the project manager informed the practice of whether he would take any specific steps in response to the particular incident (such as contacting pharmacies, IT companies, or pharmaceutical companies). If the incident was ascribed to a hospital, the report was forwarded to the existing national and regional set-up for dealing with incidents related to the hospitals
Participation and types of incidents reported	In total, the database received 422 reports during the project. Most incidents concerned matters of medication, vaccinations, and communication/cooperation in patient transitions between providers. Due to problems with the website software, it was only possible to identify the reporting unit for half of the reports received. However, it is assumed that most reports were sent in by the 45 providers (40 GP clinics, a nursing home, a pharmacy, and three hospitals) identified in the database

Table II.

Sample distribution (12 GPs from nine clinics).

Category	Variables	n
Gender	Female	3
	Male	9
Age	40–49	3
	50–60	7
	> 60	2
Practice type	Solo	2
	Group	7
Practice location	City	4
	Town	3
	Village	2

For the purpose of exploring reflections and experiences in relation to a specific system for incident reporting, respondents were recruited among the clinics that had expressed an interest in using the system by acquiring a password to the website, which was a requirement in the early part of the project. At the time of recruitment this group consisted of 29 clinics (later another 11 clinics enlisted for the reporting project). An invitation to participate in the study was distributed to the 29 clinics (via the project manager since the identity of the reporting clinics was not to be disclosed without their consent), and those who responded positively were subsequently contacted for an interview appointment. So, while a criterion of purposefulness was initially applied to narrow down a relevant study population, convenience sampling had to be employed for the recruitment of specific respondents within this population [14]. Recruitment of respondents was ended as data saturation set in, i.e. as new themes or explanations no longer emerged from the interviews [14]. (In addition to the interviews with the GPs, individual interviews were also conducted with the project manager and a representative from the regional office for quality development in order to obtain information on the background and organization of the project.)

Interview guide and analysis

An initial literature review showed that research on patient safety incidents in general practice have mainly focused on constructing and testing taxonomies and methods for describing, detecting, and counting the number and types of incidents [15–21]. Less attention has been given to issues of implementation and use. However, the studies on GP attitudes to incident reporting [3,4], as well as studies on the challenges of implementing incident reporting in the hospital sector [22–24], did suggest a number of potentially important themes, which were included in the interview guide (Table III). All interviews were audio-recorded and transcribed for analysis focusing on the major themes of the interview guide (and with an eye for emerging issues). In this phase we repeatedly went through our material comparing, grouping together, and discussing the data in order to form the central categories important for explaining the phenomenon of low use [25]. During the analysis, literature on technology acceptance [26,27] was also consulted for inspiration. Although this research strand traditionally applies quantitative methods for predictive purposes, the key notions of users’ “perceived ease of use” (or effort) and “perceived usefulness” [26,27] helped to structure the analysis.

Table III.

Themes of the interview guide.

1	Initial reflections when invited to participate in the incident reporting project.
2	The actual use of – and experiences with – the reporting system (how many reports had been submitted, who had done the reporting, how had the effort of reporting been experienced, what type of incidents had been reported, what kind of feedback had been received, what kind of changes if any had resulted from the process)
3	Reasons for reporting or not reporting specific patient safety incidents (here we asked the GPs to describe and reflect on instances in which they had reported an incident, as well as instances in which they had experienced incidents that they had not reported)
4	Potential concerns of exposure, blame, or sanctions in regards to reporting
5	The role of institutional pressures or incentives to report patient safety incidents

Results

Low reporting activity in spite of initial interest

When the project manager first set out to promote the project he was well received by the GPs in this study who expressed an interest in reporting patient safety incidents. In the interviews the GPs also conveyed several positive attitudes regarding the idea of incident reporting:

In the beginning I was very excited about this…. For me, this … is about changing inexpedient behavior so that it can be avoided in the future. (GP1)

It sounded exciting and I wanted to be a part of it…. It's important to report these incidents to learn from them and improve patient safety, and it's an area that has been neglected. (GP6)

However, in spite of this initial interest in the reporting system and the stated intentions of adoption, it turned out that the respondents generally reported few incidents, and most of the GPs declared they could have reported a lot more incidents than they actually had done. At the time of the interviews (> 1 year after project start-up) two clinics had not reported any incidents apart from those related to the promotional visit of the project manager, four clinics had reported 2–5 incidents, and two clinics had reported approximately 10 incidents while one large clinic had reported approximately 20 incidents. The latter was also the only clinic in which incident reporting had become an issue involving staff as well as GPs. The reporting pattern found among our respondents was estimated as being fairly representative of the population of GP clinics participating in the project.

Perceptions of and attitudes towards the effort of reporting

The respondents found that the project leader had done a good job at introducing them to the system by providing hands-on instruction in front of the PC, and they also perceived the electronic reporting form to be simple and easy to comprehend. Nevertheless, the GPs’ experience was that making a report could “easily” take 15 minutes due to the work of remembering, gathering information, and describing the incident in writing:

In the evening when the kids have gone to bed, you're answering emails from patients and looking at comments from the hospital and sending reminders … so to take out half an hour or so to look back in a record and bring out the story to be told …. You have to tell the story and think about what happened, so it's not just five minutes. (GP3)

The reporting form is easy to deal with … but you do have to find some data first and try to remember what happened, and so it may easily take fifteen minutes to make a report (GP11)

Such an allocation of time was especially considered problematic because incident reporting was generally experienced as a somewhat tedious administrative task taking away time from clinical work:

Reporting requires you to go out of the [current] patient [record] and do something other than treating patients and that means that it swallows up time…. My available time is to be used on my patients. It's not administration that I first and foremost should do, and that's probably the reason why I have not been as attentive as I perhaps should have been, because I find it less important than treating patients. (GP12)

We haven't put in a great effort here … we could do it, but it is boring [to report] …. I am usually able to dedicate myself to such things, but on this matter I can't whip up much enthusiasm. (GP5)

We have a working day with four patients an hour and then in the afternoon you're honestly too exhausted [to do a report] so you go home thinking “I will do it next time” … there are so many changes in our system that we have to relate to every day …. (GP1)

Thus, reporting was usually not dealt with during clinical working hours, but continually postponed and this postponing itself contributed to making reporting an unfamiliar task that required an extra amount of energy and motivation to engage in:

GP8: …I have not been looking at it [for some time]

GP9: [jokingly] It's like those stupid death certificates. If only more people were dying…. [laughing]

I: So that you would get more accustomed to use the system?

GP9: Yes, because as time passes on I can't remember how to do it….

GP7: Yes, that's right. It also takes time to locate it [the online reporting form].

Again, such an analogy to death certificates (or pharmaceutical side effect reporting, “which we don't do either” as another GP jokingly commented) signified a view of incident reporting as an unexciting administrative activity. The fact that the clinics were not remunerated for participating in the project (in contrast to previous pilot-projects in Danish general practice) did “not increase motivation” to report systematically either, according to some respondents, although others believed that incident reporting should not be turned into a financial issue.

Perceptions of usefulness

Generally, the GPs did not consider formal incident reporting to have a considerable potential for improving working processes and patient safety in the clinic. Some of the GPs believed that the learning potential of most incidents in general practice primarily applied to the clinics directly involved, which made the incidents less relevant to report:

When something originates here in the clinic it's easier for me to use it, because the consequences have to be decided upon in-house. The problem is right here, so it is here we have to act in order to prevent it from happening again. (GP2)

I am having trouble in seeing the actual learning in this, what I should use it for…. Over the years I have learned to deal with these things so that I don't make huge mistakes, so [today] it's mostly a matter of minor practical mistakes … and I can't see how anyone else can learn a lot from these apart from the fact that you have to watch out…. (GP12)

While formal reporting, analysis, and feedback was in principle seen as a relevant way of creating awareness of potential areas of risk, the respondents could not provide examples of incidents from other clinics (disseminated in the newsletters) that had given rise to specific changes in the respondents’ own clinics. However, some of the respondents had received feedback from the project manager that their reports had triggered certain changes at the local hospital. And almost all of the respondents pointed to a case in which a potentially problematic vaccination unit was improved on by the pharmaceutical company as a result of a process initiated by a report from one of the clinics. Nevertheless, such examples of usefulness were apparently not sufficient to motivate the GPs to spend more time on reporting incidents systematically. Thus, usefulness was also a relative issue in the clinic in two ways: First, although most of the GPs were positive towards the idea of incident reporting, other activities could be regarded as more useful for improving quality and patient safety in the clinic:

… within these walls there are hundreds of things that one could take an interest in and spend time on …. One of the things I am trying to implement in this clinic is diagnosis coding, which I think is a useful tool … I have spent a lot of time on registering my diabetes patients [in the national database] and that works really well. I find more quality development in doing that compared to incident reporting…. (GP2)

The areas of improvement that we have focused on here in the clinic have … mainly been about medication…. We employ targeted efforts in areas that we know are in need of a brush up. And that has not been due to incident reporting, unfortunately [laughs]. (GP5)

Second, the usefulness of reporting was relative to the specific situation and type of incident. Thus incidents occurring when communicating with the hospitals sometimes required immediate action from the GP, e.g. calling the ward to inform the relevant person of the problem so that the necessary actions could be taken to help the patient. And when an incident had already been informally brought to the attention of the hospital staff, the perceived usefulness of making a formal report decreased. In such cases, reporting was sometimes viewed as redundant and sometimes as inappropriate (cf. the next section).

Reporting as a sensitive issue in relation to other professionals

Generally the GPs did not express strong concerns about exposing themselves when reporting an incident in which they had been involved. And nearly all of the GPs were aware that the reporting system was not aimed at sanctioning and was separated from the parallel scheme for complaints over medical errors. So although a few GPs originally had some concerns over exposure this was not presented as a reason for not reporting. However, sending in reports on other health professionals was regarded by some respondents as a somewhat formalistic and harsh step – especially if the incident had already been brought to the attention of the professionals involved. Regarding an incident in the clinic, one GP recounted the following:

In a way it's a sensitive issue. We had a case that we did not report. It was an elderly patient whose creatinine had doubled without the nurse noticing it…. He ended up in dialyses but survived. It was a serious incident. My partner had a talk with the nurse about what happened and afterwards I asked him [the partner] if we should report it. “Well …” he answered, and I did not press the issue further …. Filing a report may brand the nurse and apparently he did not pursue that confrontation. (GP3)

This rather cautious approach could also apply to incidents involving colleagues at the hospitals. Here a GP comments on a situation in which he had discovered that the hospital had been much too slow in processing the X-ray pictures for a patient with the suspicion of cancer:

I do not report such incidents, because I contacted the doctor in question and he took full responsibility and did not try to explain away what had happened…. Since things were dealt with in this way I did not feel like taking the case through the official channels. Then this colleague would experience that “here it is again” … I wouldn't do that. If I can contact a given person directly and say “this is not good enough”, I prefer to do so rather than starting up this machinery. (GP2)

However, such considerations seemed to be more pronounced when the GP had a good working relationship with the hospital. Thus, a few GPs had used the reporting system to call attention to aspects of hospital behavior that they did not find to be acceptable, such as waiting times for X-rays and issues of referral.

Discussion and conclusion

While previous studies [3,4] have reported positive attitudes to incident reporting among Danish GPs, the results of this study suggest that formal reporting of patient safety incidents may be more attractive to GPs as an idea than as an integrated (non-remunerated) activity in a busy clinic where several concerns and tasks compete for attention. Thus, while most GPs expressed sympathy toward the concept of documenting and learning from patient safety incidents, actual reporting activity was generally low. In practice the GPs did not prioritize systematic incident reporting as a tool to improve patient safety, and for several GPs reporting represented a formalization of professional relations, which was often not found to be worthwhile and/or appropriate in the situation at hand.

A number of features make the particular constellation of project/respondents interesting as a case for studying the reasons for not reporting patient safety incidents in general practice. First, the project was operated by a local GP, and GP ownership of such reporting systems can be expected to lower concerns over exposure and misuse of data. Second, the project manager as a local and respected GP was in a favorable position to promote and demonstrate the system at meetings with colleagues. Third, the GPs in this study took an early interest in the reporting project, and most of them were actively involved in quality development issues, holding various positions as advisers/coordinators on cooperative relationships between general practice and hospitals and municipalities; and these GPs were already acquainted with the concept of incident reporting when the project manager first contacted them. Hence, the case can be considered to constitute a “critical case” [28] in the sense that the reasons for not reporting patient safety incidents among the GPs in this study are quite likely to be found elsewhere in general practice. On the other hand, since this study intentionally focused on relatively interested doctors, additional reasons for not reporting patient safety incidents may be found in the wider population of GPs. Still, and although the number of respondents in the study may seem small, we contend that the critical nature of the case allows for generalization on two points: (i) There is a considerable gap between attitudes and actions regarding incident reporting in Danish general practice; (ii) There are a number of important reasons for not reporting patient safety incidents in general practice – reasons that are likely to cause substantial problems for the legally supported visions of systematic and comprehensive reporting of (and learning from) patient safety incidents in primary care. Hence, the results of the study are not particularly elevating for actors promoting incident reporting in general practice. For although individuals and organizations may over time change their perceptions and enactments of what is easy and useful [29,30], the results indicate that the institutionalization of incident reporting in general practice will by no means happen swiftly and “automatically”, but will require a determined and continuous effort on the part of the regions responsible for implementing the national reporting system. Here, much will depend on the various strategies and resources employed to establish regional organizational structures for managing ongoing promotion, support, and feedback. Consequently, future research in this area should investigate how specific ways of implementing and organizing incident reporting in the various regions influence local activities of reporting and learning in general practice.

Funding

The study was funded by the Danish Research Foundation for General Practice.

Conflicts of interest

The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.