Identifying patient‐centred recommendations for improving patient safety in General Practices in England: a qualitative content analysis of free‐text responses using the Patient Reported Experiences and Outcomes of Safety in Primary Care (PREOS‐PC) questionnaire

Abstract Background There is a growing interest in identifying strategies to achieve safer primary health‐care provision. However, most of the research conducted so far in this area relies on information supplied by health‐care providers, and limited attention has been paid to patients’ perspectives. Objective To explore patients’ experiences and perceptions of patient safety in English general practices with the aim of eliciting patient‐centred recommendations for improving patient safety. Methods The Patient Reported Experiences and Outcomes of Safety in Primary Care questionnaire was sent to a random sample of 6736 primary care users registered in 45 English practices. We conducted a qualitative content analysis of responses to seven open‐ended items addressing patients’ experiences of safety problems, lessons learnt as a result of such experiences and recommendations for safer health care. Results A total of 1244 (18.4%) participants returned completed questionnaires. Of those, 678 (54.5%) responded to at least one open‐ended question. Two main themes emerged as follows: (i) experiences of safety problems and (ii) good practices and recommendations to improve patient safety in primary care. Most frequent experiences of safety problems were related to appointments, coordination between providers, tests, medication and diagnosis. Patients’ responses to these problems included increased patient activation (eg speaking up about concerns with their health care) and avoidance of unnecessary health care. Recommendations for safer health care included improvements in patient‐centred communication, continuity of care, timely appointments, technical quality of care, active monitoring, teamwork, health records and practice environment. Conclusion This study identified a number of patient‐centred recommendations for improving patient safety in English general practices.


| INTRODUCTION
Patient safety, defined by the World Health Organization as "the prevention of errors and adverse effects to patients associated with health care," 1 is a clear priority for most health-care systems. 2 Research on patient safety has been largely centred around hospitals, 3 and primary care has been perceived as a low technology environment where safety would not be a problem. However, in England, 90 per cent of contacts with the National Health Service take place in primary care, and more than 750 000 patients consult their GP each day. 4 A recent systematic review including studies from 21 different countries 5 estimated that 2-3 patient safety incidents occur per 100 primary care consultations, and 4% of them result in severe harm (long-term physical or psychological problems or death). In the UK, this would translate to between 15 000 and 22 500 safety incidents per day, resulting in between 600 and 900 patients being severely harmed each day.
Between 45% and 76% of these incidents could be prevented. 6 Most of the research conducted so far in the area of primary care patient safety is based on information supplied by health-care providers, 7 and limited attention has been paid to patients' perspectives. [8][9][10] Patients are the common element across the various settings, organizations and health professionals usually involved in their health care, and therefore, they are ideally suited to reflect on the health care they receive. As highlighted by World Health Organization in a recent report, 11 tapping into such a rich resource could contribute significantly to improving safety in primary care.
A number of recent qualitative studies have examined patients' perceptions of different aspects of patient safety in primary care, including the ways in which patients make sense of "safety" in the context of primary medical care, 12 their perceptions of errors in long-term illness care, 13 the effect of workplace conditions on errors, 14 what they believe may be done to reduce errors, [15][16][17][18] and how safety problems may impact on their subsequent interactions with the health-care system. 19 Although important progress has been made in this area during the last ten years, this is a relatively new field and further research is needed to better understand patients' perceptions and experiences of safety problems in English general practice. Previous studies are heterogeneous in terms of the different aspects of patient safety examined, but also in terms of countries in which they have been conducted (Australia, New Zealand, USA) with diverse health systems. Patient safety is highly contextual, and findings cannot be necessarily extrapolated across countries. The available evidence in the UK (a country with strong primary care orientation) is still scarce, with only four studies currently published. 12,13,17,20 Also, previous qualitative studies relied on data obtained through focus groups or individual interviews, including a relatively low number of participants.
Additional qualitative research using alternative methodological approaches (eg qualitative content analysis of free-text responses to a survey completed by a large number of participants) may contribute to a better understanding of patients' perceptions and experiences of patient safety in primary care.
The aim of this work was twofold: (i) to explore patients' perceptions and experiences of patient safety in general practices in England and (ii) to identify patient-centred recommendations to improve patient safety in primary care.

| Questionnaire
Data were collected with the Patient Reported Experiences and Outcomes of Safety in Primary Care (PREOS-PC) questionnaire. 21 PREOS-PC was developed in a multistage process supported by an expert panel and informed by two systematic reviews, 22,23 four focus groups, 17 18 cognitive interviews and a pilot study which involved 1975 patients registered in 26 general practices. 21 Available evidence supports the validity and reliability of the questionnaire. 21 The standardized items in the survey measure different domains of patient safety including patient and practice activation for patient safety, experiences of safety problems, harm and overall perceptions of patient safety.
In addition, PREOS-PC included seven open questions. Four of them asked patients about their experiences of safety problems and harm, whereas the other three asked about lessons learnt as a result of experiencing a safety problem; good practices followed by health-care professionals to ensure the provision of safe health care; and suggested changes to improve patient safety in their practices (see Table 1).

| Data collection
In June 2014, the questionnaire was sent to 6736 adult (18 years old or older) patients from 45 general practices distributed across five regions in the north, centre and south of England. Practices were selected using purposive sampling to ensure variation in terms of list size and levels of deprivation. 21

| Analysis
Firstly, data were cleaned by removing free-text responses that contained no relevant information eg "N/A" or "No comments." Clean data were then analysed using conventional content analysis. 26 A qualitative researcher (LSC) read all data repeatedly to get a clear understanding of the entire dataset 27 and then coded these data 28,29 by first highlighting the exact words from the text that appeared to capture key thoughts or concepts. The researcher also made notes of her first impressions and initial analysis. The coding scheme was developed inductively from these data, with codes either coming directly from the text or reflective of one or more key thought(s).
An inductive approach was followed because no suitable theoretical framework to test or explore was identified. Key themes were identified in the answers to the open-ended questions, which were often related. The preliminary coding scheme was discussed with a second researcher (IRC) and revised. All data within each code were re-examined by the two researchers. Codes then were sorted into categories based on how different codes were related and linked.
These emergent categories were used to organize and group codes into meaningful clusters. 30 All analyses were conducted separately for each question, except for questions 6 and 7 (which were combined because of their substantial overlap in the underlying question). Throughout the analysis process, a third analyst (JMV) was involved for triangulation purposes. A limited number of direct quotes from participants have been used to convey some important themes. Data were analysed using the NVivo 10, a data management and analysis software.

| Response rate
The overall response rate to the PREOS-PC questionnaire was 18 Two main themes were identified (i) experiences of safety problems and harm and (ii) good practices and recommendations to improve patient safety in primary care ( Figure 1).

| Experiences of safety problems and harm
Two subthemes were identified for experiences of safety problems and harm: types of safety problems and harms experienced, and patients' responses after experiencing a safety problem. Educational attainment (% with degree and above)

Health status (% good/ very good)
Long-term conditions (% one or more long-term conditions) 1. In case you experienced more than one safety problem in the last 12 months, which of the following better describes the most recent safety problem you experienced? Please select all the boxes that apply to you. Please feel free to describe here in more detail the most recent problem that happened to you. were related to the appointment system (n=117), followed by the coordination between health-care providers (n=46), diagnostic tests (n=41), medication (n=35) and diagnosis (n=32). Harm was generally described in terms of emotional harm (eg anxiety, stress, concern or panic attacks), which were generally produced as a result of delays in obtaining an appointment or in receiving a diagnosis or adequate treatment. Some participants reported that these delays resulted in their condition being unnecessarily extended or exacerbated. In some instances, the harm experienced affected not only the patients but also their families or friends.

| Patients' responses to experiencing a safety problem
A total of 181 participants provided information on their responses after experiencing a safety problem in the surgery. Becoming more active players in their own health care (increased patient activation) 21 emerged as the most important response (Box 2). This mainly involved speaking up about concerns they may have had in relation to their symptoms or given diagnosis. Some participants also reported seeking a second opinion when they perceived their problems were not being taken seriously by their GPs, double-checking the accuracy of the information in their health records or prescriptions and proactively requesting test results or more timely appointments. In addition to this increased proactivity, the avoidance of unnecessary exposure to health care emerged as an additional response to a safety problem experience.

| Good practices and recommendations to improve patient safety in primary care
A wide range of factors perceived to mitigate the occurrence of safety problems and harm in general practices emerged from participants' responses to the questions about "good practices" for safe care (452 respondents) and about "suggestions" to improve safety (422 respondents). Box 3 outlines the main subthemes identified.
Timely appointments emerged as an important component of safe health care, being identified both as a good practice and a suggestion Problems with the appointment system included difficulties to get an appointment when needed or when convenient, and difficulties to get an appointment with their preferred GP.
"Availability of appointments is the main problem (…). The on-line system doesn't work well, neither does booking on the automated telephone system. We are therefore having to ring at exactly 8AM and wait for what seems ages for the phone to be answered and then might be offered an appointment for the following week." (male, 60 years) Coordination between health-care providers (N=46) Patients reported miscommunication and lack of cohesion among staff in the practice and between levels of care. Some participants reported significant delays in being referred to a specialist, which resulted in delayed diagnosis and treatment.

Medication (N=35)
Medication-related problems included prescribing the wrong type or dose of medication or treatment duration. Patients attributed these problems to clinical or administrative mistakes. They also reported experiences of adverse drug reactions and of GPs expressing reluctance to change their medication plan.

Communication (N=29)
Lack of patient-centred communication (eg GPs not listening or believing to their patients, or not explaining to them important aspects concerning their condition, treatment or prognosis) resulted in delayed diagnosis or referrals, which in some instances caused emotional harm to patients.
In addition, confidentiality of information was not ensured by the staff working in the reception area.

Health records (N=21)
Health records containing outdated or wrong information caused a delay in diagnosis, prescriptions and reviews. Some patients reported that their health records were not available when needed, which was perceived to be due to fails in transferring the information between levels of care, lack of electronic information and inefficiency in the process of updating records.

"My Drs never has hospital letters available to read as "they take a while to scan in the computer and put on your notes"-This is what I'm told even 3 weeks after they have received the letter (I get a copy at the same time) I have to take my copy in and get them to
scan it-Not sure why!" (female, 33 years)

Box 1 (Continued)
Box 3 Good practices and recommendations to improve patient safety in general practices Access and appointments (N=209) "You are able to telephone in if you have an urgent medical problem a nurse will call you back, speak to you, and then will arrange for you to see a doctor that day if she feels the need is urgent, excellent service" (female, 75 years-observed good practice).
"Improve on-line and automated systems. Penalise more those who fail to attend appointments" (male, 60 years-suggestions to improve safety).

"Open to longer surgery times plus open on weekends.
More doctors needed to reduce waiting times" (female, 59 years-suggestions to improve safety).

Patient-centred care(N=66) "My GP listens very well to me when I share concerns of my condition. Always is happy to discuss medication + treatment. Allows and helps me to feel very involved in my care and also takes seriously how I fell" (female, 40 years-observed good practice).
"Doctors to listen to patients when speaking-not he reading computer screens." (female, 71 years-suggestion to improve safety).

Active monitoring (N=64)
"Regular blood tests, diabetic clinic, practice concerns itself more about my wellbeing than I think I do" (male, 83 years-observed good practice).

"The only thing I could suggest is re: feedback from blood tests etc. it would be good if surgery could ring up and say 'all clear' rather
than having to assume all is well because you have heard nothing" (female, 62 years-suggestion to improve safety).
"Always follow up important/life changing/emergency appointments" (female, 54 years-suggestion to improve safety).
Training and technical quality of clinical care (N=50)

"Evidence based, high standard of protectional care delivered in a friendly setting by a very helpful team of GPs/Nurses and other
staff" (male, 54 years-observed good practice).
"Junior doctors [to be] supervised more. I think they have done years of training before they come to our surgery but they still need more supervision" (female, 41 years-suggestion to improve safety).

Teamwork (N=45) "I believe communication between all members of my practice is excellent, leaving me feeling 'in good hands' from reception, doctor
to pharmacy." (male, 71 years-observed good practice).
"More conversation and control over the procedures we are sent for in other places." (female, 48 years-suggestion to improve safety).

Environment and equipment (N=37)
"Sanitiser for hands supplied when you walk into the surgery." (female, 59 years-observed good practice).
"Make sure equipment is working at all times" (female, 43 years-suggestion to improve safety).
(Continues) for improvement. Participants highlighted the importance of being able to book same-day appointments when they were ill. They found inefficient those appointment systems operating on a "first ring, first "I would suggest that GPs always check the notes of the patient they are seeing and that they listen to what the patient is describing fully" (female, 44 years-suggestion to improve safety).

Continuity of care (N=19)
"Provide access to same doctor on return visits" (male, 56 years-suggestion to improve safety).
"To be able to see the same doctor. Then they know your history more than just having a quick glance at medical history" (female, 64 years-suggestion to improve safety).
"Not so many part-time and short stay doctor-very difficult to see your own doctor, only works part time and away on quite a few holidays" (female, 71 years-suggestion to improve safety).
Seek patients' feedback (N=4) "Perhaps carry out at more frequent surveys with customers such as this one. This is the first I have completed, I think is over 30 years attending the surgery." (female, 59 years-suggestion to improve safety).
"Make complaints procedure more obvious and accessible." (female, 45 years-suggestion to improve safety).

Box 3 (Continued)
information. Participants also mentioned the importance of pharmacies in double-checking the accuracy of their health-care records when dispensing medication, and underscored the positive role of IT system in this issue.
Continuity of care emerged as an important factor related to patient safety. Participants perceived their usual doctors to be more familiar with their own medical history and therefore less prone to diagnosis or treatment-related errors. They suggested that in those cases where continuity of care cannot be offered, GPs should carefully review patients' health records prior to the consultation.
Some participants suggested that a structured engagement and feedback from patients should be encouraged to achieve safer health care.

| DISCUSSION
This qualitative study examined patients' perceptions and experi-

| Strengths and limitations
In contrast with previous qualitative studies on primary care patient safety, this study is based on data from a large number of participants from 45 practices distributed across England. The study also builds on and expands previous related research focusing on patients' perceptions of safety in primary care 12,13,16,17,20,22,31,32

| Comparison with previous literature
The type and nature of safety problems and harm experienced by patients are generally in line with previous qualitative work 13,16,17,32 and with the quantitative data from the PREOS-PC survey. 25  Continuity of care also emerged as an important attribute of patient safety in primary care, which supports results from previous research. 17,20 Ensuring that patients are able to see the same GP enables the GP to become a repository of information; acquire specialist knowledge of a patient's condition; become familiar with the patient's consulting behaviour; provide holistic care; and foster the development of trust. 45

| Practice implications
Health-care professionals and commissioners of English general practices should be aware of the factors that patients perceive can influence safer health care. Even though patients' perceptions of safety problems may not always result in adverse events, they could, however, influence patient satisfaction, which has been associated with a higher engagement of health services and increased treatment adherence. 46,47 Practices should therefore consider implementing evidence-based strategies to improve patient perceived safety. 11 Although the evidence base of interventions to improve safety in general practices is still scarce, most of the factors identified by patients were specific and fall within the realm of health-care quality, where available evidence is stronger. 2 For example, evidence suggests that GP or nurse-led telephone triage could be effective to improve access to same-day consultations, 48 which is one of the most frequent recommendation from patients to achieve safer health-care delivery.
Patients made a large number of recommendations to improve different areas of patient safety in general practices. Research is now needed to explore the acceptability and perceived utility of those recommendations by health-care professionals and commissioners; to identify effective strategies to support their implementation in a context of resource limited service; and to measure its impact.
Finally, practices may be heterogeneous in terms of the areas they need to improve to deliver safer health care (eg, some practices may be perceived to provide excellent patient-centred care, but struggle to offer timely appointments, or vice versa). The use of standardized and validated patient reported instruments, such as the

Patient Reported Experiences and Outcomes of Safety in Primary
Care (PREOS-PC) questionnaire, 10 might prove a valuable resource for practices in order to help them identify and prioritize areas for safety improvement.

| CONCLUSION
This study identified a number of key areas that patients believed influenced the safety of health care provided in their general practicesnamely access to appointments, quality of clinical care, psychosocial relationship with health-care providers or continuity of care-and helped us increase our understanding of patients' behavioural responses to experiences of safety problems and harm. The information gathered in the open-ended questions complemented the quantitative data generated from the standardized items in the questionnaire, 25 allowing us to better understand specific aspects of patients' experiences and perceptions of safety problems in general practices in England.