Challenges and Strategies for Patient Safety in Primary Care: A Qualitative Study

Abstract

Although most health care occurs in the ambulatory setting, limited research examines how providers and patients think about and enact ambulatory patient safety. This multi-method qualitative study seeks to identify perceived challenges and strategies to improve ambulatory safety from the perspectives of clinicians, staff, and patients. Data included interviews (N = 101), focus groups (N = 65), and observations of safety processes (N = 79) collected from ten patient-centered medical homes. Key safety issues included the lack of interoperability among health information systems (HIT), clinician-patient communication failures, and challenges with medication reconciliation. Commonly cited safety strategies leveraged HIT or involved dedicated resources (e.g., providing access to social workers). Patients also identified strategies not mentioned by clinicians, emphasizing the need for their involvement in developing safety solutions. This work provides insight into safety issues of greatest concern to clinicians, staff, and patients and strategies to improve safety in the ambulatory setting.

INTRODUCTION

Disproportionately little research has addressed patient safety in primary care, although most health care is delivered in this setting.¹ A growing literature has identified specific areas of concern for ambulatory safety, including hospital transitions, medications, and health information technology, as well as corresponding strategies to improve safety, such as ambulatory follow-up after transitions, medication reconciliation, and e-prescribing.^2,3 However, key gaps in knowledge about ambulatory safety remain, particularly around diagnostic safety,⁴ challenges faced by patients and caregivers,⁵ and the impact of restructuring primary care on safety-related processes and outcomes.²

The patient-centered medical home (PCMH) model, which aims to provide care that is comprehensive, coordinated, patient-centered, accessible, and safe,⁶ is a model of care that intersects with patient safety in important ways. Many PCMH standards have the potential to improve safety, such as medication reconciliation for medication safety, and care coordination to address delays in diagnosis and treatment.⁷ And yet, in complex adaptive systems such as healthcare, ‘work-as-done’ by frontline staff and clinicians is often different from ‘work-as-imagined’ by those who develop guidelines and protocols.⁸ Consequently, the perspectives of frontline staff and patients are critical to understanding how safety is actually conceptualized and enacted in the ambulatory setting.

This study expands our knowledge of ambulatory patient safety by identifying perceived safety issues and strategies to improve ambulatory safety. It does so by integrating clinician, staff, and patient perspectives as well as observations of key clinic processes related to safety in a diverse sample of PCMHs. Previous research has explored participants’ views of the meaning of ambulatory safety⁹ and factors thought to contribute to patient safety incidents.¹⁰ However, these reports have not examined participants’ views of: (1) the most salient safety issues in the ambulatory setting; and (2) what strategies individuals and organizations use to address these challenges. Such information can be used by policy makers and practitioners to design and prioritize policies and practices that incorporate what frontline staff and patients believe are the most critical aspects of ambulatory safety.

METHODS

Conceptual Framework

This study employed the positive deviance approach for studying “what works” in organizations with high performance¹¹ to understand ambulatory safety in a potential best-case scenario. Data collection and analyses were guided by an ambulatory safety framework based on two expert-consensus projects, a systematic review, and an analysis of incident reports.^2,3,12,13 The framework includes four key safety domains: (1) missed, delayed, or incorrect diagnoses; (2) delays in treatment or preventive services; (3) medication safety issues; and (4) issues of communication and coordination of care.

Study Design and Setting

This is a qualitative study of 10 PCMHs in four US states (Pennsylvania, Maryland, North Carolina, and Colorado). Consistent with the positive deviance approach, all the PCMHs had achieved Level 3 NCQA recognition status. Purposive sampling was used to recruit PCMH practices that varied by practice type, population served, and geographic location (Table 1).

Table 1.

Site Characteristics

PCMH Site Characteristics	Value
Practice type, no. of sites
Federally Qualified Health Center	1
Hospital/health system-owned	5
Independent clinician-owned	4
Population predominantly served, no. of sites
Rural	3
Suburban	6
Urban	1
Location, no. of sites
Colorado	2
Maryland	1
North Carolina	3
Pennsylvania	4
Median no. of years since becoming Level 3 PCMH at the time of data collection, (range)	3.5 (1-7)
Median patient panel size, by site (range)	5,700 (120-32,000)

Data Collection

Two-day site visits were conducted at each site from May 2017 to April 2018. The site visits included: (1) in-depth interviews or focus groups with clinicians and staff (N = 101 participants); (2) focus groups with patients and caregivers (N = 65 participants); and (3) in-situ observations of PCMH staff as they conduct routine activities related to patient safety (N = 79 participants) (Tables 2 and 3).

Table 2.

Clinician and staff characteristics

Demographics / Study participants	Interview Participants	PCMH Staff Observed*
Profession, no. of participants (%)
Administration or management	37 (36.6)	3 (9.6)
Medical receptionist		2 (6.4)
Care coordinator/navigator/patient access rep	9 (8.9)	3 (9.6)
Medical assistant	11 (10.9)	16 (51.6)
Nurse	8 (7.9)	1 (3.2)
Physician	25 (24.8)	1 (3.2)
Other (dental staff, pharmacist, physician assistant, social worker, behavioral health specialist, support staff, intern)	11 (10.9)	5 (16.1)
Median no. of years working at site, by participant (range)	4 (1 month–29 years)	6 (<1 year–35 years)
Median no. of years working in healthcare, by participant (range)	15 (<1 year–43 years)	8 (<1 year–35 years)
Sex, no. of participants (%)
Female	78 (77.2)	30 (96.7)
Male	23 (22.8)	1 (3.2)
Race, no. of participants (%)
Asian	1 (1)	0 (0)
Black/African American	3 (3)	1 (3.2)
Native American	1 (1)	0 (0)
White	86 (85.1)	29 (93.5)
More than one race	1 (1)	1 (3.2)
Other	1 (1)	0 (0)
Did not wish to answer	8 (7.9)	0 (0)

^*Note: Six staff members were observed more than once

Table 3.

Patient characteristics

Demographics / Study participants	Focus Group Participants	PCMH Patients Observed
Sex, no. of participants (%)
Female	35 (53.8)	30 (71.4)
Male	26 (40)	12 (28.5)
Did not wish to answer	4 (6.2)	0 (0)
Race, no. of participants (%)
Asian	1 (1.5)	0 (0)
Black/African American	3 (4.6)	5 (11.9)
Native American	1 (1.5)	0 (0)
White	55 (84.6)	25 (59.5)
More than one race	0 (0)	7 (16.6)
Other	0 (0)	1 (2.3)
Did not wish to answer	5 (7.7)	4 (9.5)
Education, no. of participants (%)
Elementary/Middle school	0 (0)	2 (4.7)
High school graduate/some high school	25 (38.5)	10 (23.8)
Some college/College degree	31 (47.7)	14 (33.3)
Degree after college	8 (12.3)	8 (19)
Did not wish to answer	1 (1.5)	8 (19)
Age range, no. of participants (%)
18-45	10 (15.4)	11 (26.1)
46-65	22 (33.8)	10 (23.8)
66-75	20 (30.8)	10 (23.8)
>75	12 (18.5)	6 (14.3)
Did not wish to answer	1 (1.5)	5 (11.9)

The site visit teams included two to four members from Johns Hopkins University and one from the National Committee on Quality Assurance. The teams had extensive expertise in primary care, patient safety, organizational science, human factors engineering, and qualitative research and underwent training on the study materials and procedures.

For interviews and focus groups, semi-structured discussion guides addressed participants’ perceptions of the meaning of patient safety, functioning of the PCMH model related to patient safety, and perceived challenges and improvement strategies in the four domains (Appendix Tables 1 and 2). Interviews lasted approximately 30 – 60 minutes and focus groups 60 – 90 minutes. Both interviews and focus groups were audio-recorded and professionally transcribed.

Observations used contextual inquiry, a process of asking questions of those performing tasks while tasks are underway or immediately after,^14,15 to explore key workflows and processes related to patient safety. Observation periods, which lasted approximately 60 minutes each, consisted of two research team members shadowing a patient during each stage of an office visit (e.g., check-in, medication reconciliation), excluding the physician-patient encounter. The researchers recorded observations using a semi-structured observation guide (Appendix 3). Immediately following the observation period, the researchers debriefed the patient and corresponding medical assistant separately to ask about perceptions of the office visit and tasks, respectively. The Johns Hopkins Bloomberg School of Public Health Institutional Review Board (no. 7497) approved all research procedures. All participants provided written informed consent.

Data Analysis

A multidisciplinary coding team (CY, AL, EL, SH, JH, TO) identified an initial set of codes a priori using the ambulatory safety framework. The coding team then reviewed transcripts and observation notes line-by-line using the constant comparative approach¹⁶ to inductively identify new codes that emerged from the data and revised the coding structure accordingly. Disagreements in coding and interpretation were resolved through discussion and group consensus, complemented by regular debriefings with the full team. When the coding structure was considered final (i.e., no new concepts were apparent), two independent coders analyzed 20% of all materials independently to assess intercoder agreement. Coders achieved at least 80% agreement on code frequency and code existence for each transcript. When all coding was complete, the team collated codes for each domain and synthesized recurrent themes related to perceived challenges and strategies addressing patient safety. MAXQDA 12¹⁷ was used to organize data and facilitate analysis. To enhance the credibility of findings, the team used an intuitive approach to triangulate data ¹⁸ across data collection methods (interviews, focus groups, and observations) and respondent groups (clinicians, staff, and patients). Feedback on study materials and results were obtained from patient advisory and expert advisory groups.

RESULTS

Findings related to perceived safety issues (Table 4) and suggested strategies to mitigate these issues (Table 5) are summarized by the four ambulatory safety domains.

Table 4.

Overview of perceived causes of safety issues (across domains)

Domain	Top Themes	Support from Data Sources
		Clinician / Staff Interviews	Patient Focus Groups	Observation Data
Missed, delayed, or incorrect diagnoses	Lack of interoperability between health information systems*	●	●	●
	Perceptions that EHR “problem lists” included outdated / incorrect diagnoses	●		●
	Lack of time to collect and synthesize patient information	●
	Communication failures between providers and patients	●	●
Delays in treatment or preventive services	Extended wait times to see specialists, particularly in behavioral or mental health	●	●
	Lack of follow-through with recommended screenings and diagnostic procedures	●	●
	Lack of clear workflows to support health information technology (HIT)	●
Medication Safety Issues	Challenges associated with patients’ adherence to medications	●	●
	Difficulties in maintaining a complete and accurate medication list	●	●	●
	Shortcuts during the medication reconciliation process	●		●
Issues of communication and coordination of care	Lack of interoperability between health information systems*	●		●

^*Denotes a cross-cutting theme that was present in multiple domains

Table 5.

Overview of safety strategies (across domains)

Domain	Top Themes	Support from Data Sources
		Clinician / Staff Interviews	Patient Focus Groups	Observation Data
Missed, delayed, or incorrect diagnoses	Using electronic health record (EHR) systems to track test results / flag abnormal and overdue tests	●		●
	Using protocols to systematically gather new patient information, particularly after transitions of care	●		●
	Using a team-based approach during the diagnostic process	●
	Scheduling enough time for physicians to explore patients’ underlying conditions	●	●
	Providing coordinated “next steps” following a visit		●	●
Delays in treatment or preventive services	Increasing access to care through more timely appointments, on-site laboratory testing, and on-site specialty services	●	●	●
	Leveraging health information technology (HIT) systems to track outstanding referrals / health screenings	●		●
Medication Safety Issues	Having clinical pharmacists on staff	●	●
	Using HIT systems to check for drug-drug interactions	●		●
	Facilitating medication reconciliation by printing medication lists for patients during the visit / asking patients to physically bring in pill bottles	●	●	●
	Conducting medication reconciliation in the patient’s home	●
	Providing patients with physical descriptions of medications (e.g. shape, color)		●
Issues of communication and coordination of care	Appointing / hiring staff to perform care coordination activities during transitions of care	●	●
	Using the EHR’s patient portal to facilitate timely communication	●	●	●

Domain 1: Missed, delayed or incorrect diagnoses

This domain was the least emphasized across data sources and practices. For missed diagnoses, several clinicians and patients shared stories of misses and near-misses, typically stemming from a lack of diagnostic test follow-up. In contrast, delayed and incorrect diagnoses were perceived as uncommon, with both clinicians and patients recognizing the trial and error often accompanying diagnostic processes. As a medical assistant (Site 3) remarked, “I don’t know that we see any incorrect diagnoses. I imagine there probably are, but I don’t hear of that.”

Perceived causes of diagnostic safety issues fell into four key themes. First, clinicians and staff from nearly all sites experienced challenges with accessing test results from specialists or hospitals that did not share a common electronic health record (EHR). Participants noted that relying on patients and outside healthcare providers to fax information to PCMHs led to information gaps, and in turn, diagnostic issues. A second, related theme, was the perception that EHR “problem lists” were not always accurate because of outdated or incorrect diagnoses not being removed. A third reported cause stemmed from a perceived lack of time to collect and synthesize patient information, with several physicians expressing concern about reviewing and synthesizing large amounts of information before patient encounters without dedicated time. Finally, the fourth theme involved communication failures between PCMHs and patients. Several patients shared stories of misdiagnoses that they perceived resulted from physicians not listening to patients’ concerns. As one patient reflected, “the missed diagnosis I had was due to a doctor being hurried, being arrogant, and dismissive. I told him… this is in the file. He wouldn’t even look in the file” (Site 3). In contrast, several clinicians reflected that some patients were unwilling or unable to share details needed for the diagnostic process.

PCMHs used several strategies to prevent diagnostic errors. According to interviews and observations, all sites used EHRs to track test results and had systems to flag abnormal results and overdue tests. Several sites had also developed protocols to systematically gather new patient information, particularly after care transitions, to ensure problem lists were treated as “living documents” (Site 4), capturing new diagnoses, testing, or issues requiring follow-up. Clinicians and staff, particularly from sites affiliated with academic medical centers, also perceived that a team-based approach to explore differential diagnoses may improve care. Patients perceived two ways in which practices improved diagnosis processes: (1) having physicians take time to explore underlying conditions potentially causing symptoms, and (2) providing coordinated “next steps” following a visit (e.g., arranging appointments with specialists for patients).

Domain 2: Delays in treatment or preventive services

Participants largely conceptualized the second domain, encompassing delays in treatment or preventive services, in terms of access to care (e.g., offering extended clinic hours) and health maintenance (e.g., providing recommended health screenings).

For perceived causes, clinicians and staff from nearly all sites commented on long wait times to see certain specialists, leading to delays in care. This was particularly true for mental health specialists. A second challenge, echoed by both clinicians and patients, was perceived lack of follow-through with recommended screenings and diagnostic procedures. As a medical director reflected:

Prevention is tough. Obviously, you know, in primary care, we send letters. We have all these different things that we try to put in place to get people in to do their preventive care but it still comes down to: that patient still has to take action on that… you cannot make patients, other adults, do these things if they don’t want to do it. (Site 9)

A patient echoed the importance of patient responsibility, saying:

I try and help other people understand that our health is our responsibility… [I say] if you want follow through, get another appointment with your doctor. (Site 3)

A third challenge was rooted in lack of clear workflows to support use of HIT. For example, although all the clinics had applications in the EHR for test tracking, several clinicians and staff mentioned that the process of assigning staff to monitor and manage these workflows was either unclear or under-resourced, as illustrated by one nurse manager:

We’ve put a system in place into the EHR to track those things, but there’s always a resource issue. You need people to work those queues, to work those folders. So I think we’ve done a really good job as far as setting up our tools to work for us, but it always comes down to who’s working the overdue results folder this week, you know, they’re tasked with so many different things (our clinical staff) that I think we need to figure out a way to prioritize for them. (Site 4)

For preventing delays, the most widely discussed strategy was increasing access through more timely appointments (e.g., same-day appointments, extended clinic hours), on-site laboratory testing, and improved access to specialty services (e.g., having a pharmacist or social worker on staff). Leveraging the HIT system to track outstanding referrals and health screenings was also central to participants’ thinking about preventing delays and enhancing preventive services use. A physician noted using HIT for “closing loops” and “being a second brain” (Site 1), whereas others reflected on HIT use to support team-based approaches, such as daily careteam huddles to discuss care plans, that aligned with PCMH requirements. As a medical assistant reflected:

When I started, we weren’t PCMH… we didn’t look at colonoscopy or mammograms. That was doctor stuff. Now that we run our reports and everything all the time, it’s part of our huddle. We pre-visit plan. We make notes in our schedule, we discuss it with the doctor. We’re more involved. (Site 4)

Domain 3: Medication Safety Issues

Medication safety issues, including medication errors, insufficient monitoring, and preventable adverse drug events, were key to clinicians’ thinking about patient safety. As the director of quality at Site 5 remarked, “I know that medication safety issues is probably the biggest for ambulatory care. So, if you look at the list of things that cause the greatest harm, that’s it.”

Participants perceived three main causes of medication safety issues. Across sites, both clinicians and patients highlighted challenges with medication adherence, which they attributed to issues of cost and affordability (particularly given changing insurance formularies), side effects, and varying patient health literacy. For clinicians and staff, another key challenge was maintaining a complete, accurate medication list. Despite clinicians’ strong sense of responsibility for this task and perceived importance for patient safety, changes to the medication list were often difficult to track during care transitions and visits to specialists. To address this, many of the PCMHs had instituted a policy of medication reconciliation (i.e., the process of obtaining and maintaining an accurate list of all medications a patient is taking) for every patient, every visit, every single time. The frequency and routine nature of the process, however, also made it prone to staff shortcuts. For example, a physician noted:

I don’t think we do a very good job of med reconciliation. I find oftentimes that I’m backtracking, asking questions, and having to take stuff off the list or add stuff on the list and I think it’s… oftentimes… especially if the list is long, that the patient is asked a blanket ‘has anything changed’ question instead of really going through that list of medications. (Site 9)

In terms of strategies to support medication safety, several practices had clinical pharmacists on staff to conduct detailed medication reconciliations (particularly for patients with multiple comorbidities), educate patients on “when and why” medications needed to be taken, investigate insurance company formularies, check for drug-drug interactions, and simplify medication lists. Other common strategies for medication safety included using HIT to check for drug-drug interactions and printing medication lists for patients. In terms of strategies directly involving patients, some PCMHs recommended that patients bring in pill bottles for medication reconciliation or, in one practice (Site 7), had a care navigator visit patients at home for medication reconciliation. To increase medication adherence, a patient offered the following suggestion:

Because we read so many things about wrong medication, I would like to see a description [provided on the medication list] of what you’re taking… just so that when I see it, okay, yeah, I know I have the right thing. Even if it’s just through the portal… ‘your pill should look pink and be oblong and have this letter on it,’ something like that. (Site 5)

Domain 4: Issues of communication and coordination of care

The fourth domain – encompassing practice-patient communication, coordination between sites, managing transitions from hospitals to primary care, and communication of test results – was perceived to include both underlying causes and potential solutions for many of the safety issues identified in other domains. As such, this domain was the most prominent for the majority of participants. This was particularly true for patients, for whom trust and communication with clinicians were central in their perceptions about patient safety. Across safety domains, patients continually referenced the need and desire for clear communication, which they felt was enhanced by EHRs and patient portals.

For clinicians and staff, lack of HIT interoperability contributed to many communication and care coordination issues. As one physician remarked, “If we can ever get to the point where we have electronic health records talking to each other nationwide, we’re going to get rid of a lot of this. I mean, every day I’m having patients sign releases and we’re duplicating labs within the week that have been done somewhere else in town” (Site 1). Patients, on the other hand, had generally positive perceptions about information exchange between clinicians. A patient remarked (Site 5), “my husband was in the ER and was admitted to the hospital and the next day, [the doctor] had already emailed my husband to say, ‘I see you’re in the hospital. Are you okay?’ I have all of your labs and everything.”

For strategies to enhance communication and care coordination, all sites had appointed, and, in several instances, hired new staff to call patients following hospital discharge to schedule follow-up appointments, perform medication reconciliation, obtain hospital discharge records, and communicate about new medications. This strategy, and many others designed to increase care coordination, aligned with practice requirements for PCMH certification. For example, a Medical Director (Site 9) described his practice’s transformation in adopting new processes to manage care transitions to become a PCMH:

Before… a patient would show up and say I’ve been in the hospital. I mean, this could’ve been somebody who was in the hospital for four weeks and had major complications after a surgery. It was like getting War and Peace and opening it in the middle and trying to figure out what in the heck is going on with this person… so, to have those processes in place where, I mean from the front office all the way back, they know these are the protocols we’re getting a patient in after a hospital followup. (Site 9)

Patients viewed the patient portal as a valued resource enabling timely communication with the practice. As a patient reflected:

Everything’s on there. My appointment’s on there. All the medicine I take is on there. Every appointment I have in the future is on there, and if there’re any changes [they are] on there and that is, the best thing ever is to know what’s going on in your health and in your life instead of waiting. (Site 7)

DISCUSSION

Ambulatory patient safety is a critical but understudied issue. This qualitative study integrated clinician, staff, and patient reflections on what they perceived as key patient safety issues and strategies to improve safety in primary care. Findings suggest that many safety issues were rooted in communication-related challenges, both between clinicians (e.g., lack of HIT interoperability) and between patients and clinicians (e.g., clinician-patient communication failures). Lack of clear workflows, challenges with patient follow-through, and health system constraints were other notable drivers of perceived safety issues. The strategies identified by clinicians, staff, and patients tended to either increase structure (e.g., protocols to systematically gather patient information after care transitions) or access to care (e.g., having a social worker on staff). Patients also identified strategies to increase clinician-patient communication not mentioned by clinicians and staff, emphasizing the importance of their involvement in patient safety solutions.

One of the clearest perceived threats to patient safety was the lack of HIT interoperability. Given the current US HIT ecosystem, where only 14 percent of ambulatory providers share EHR information with outside providers,¹⁹ this finding emphasizes the need for improved approaches to exchange electronic health information.²⁰ Although our HIT system nationally needs improvement, many of the strategies identified (e.g., checking drug-drug interactions, using patient portals to improve timeliness of communication) were rooted in existing HIT strategies with evidence to improve patient safety.^21,22

Certain safety issues were more salient to patients (e.g., communication issues between clinicians and patients) than to clinicians and staff, who were more focused on process-oriented issues (e.g., difficulties in maintaining an accurate medication list). Issues with the diagnostic process were also not perceived to be a major issue by either clinicians or patients. Diagnostic errors often result from clinicians’ cognitive errors (e.g., not considering the correct diagnosis as a possibility),²³ which can be difficult to detect in a process largely framed as one of “trial and error.” Although there was some mention of using EHR systems and a team-based approach to increase “distributed” cognition that extends beyond a single clinician,²⁴ there was little emphasis on this domain despite the frequency of diagnostic errors in primary care.^25,26

Among the implications of this work is the need for ambulatory practices to heed the importance of the patient emphasis on communication and trust as critical elements of patient safety. The results also provide insight into other strategies to address patient safety threats, and though some strategies are more resource intensive than others, several are accessible to every practice. Finally, these findings imply an expectation for management to assess, communicate, and engage all clinicians and staff in regular assessments of critical workflows and metrics related to the safety domains. In some situations, primary care practices may need to add staff in different disciplines to meet these needs.

This study had strengths and limitations. The sample of PCMHs was diverse but limited; findings may not apply to less high-functioning sites. Second, although qualitative methods are a valuable means to understand delivery of care in real world settings, many processes and mechanisms of patient safety are difficult to observe,²⁷ and thus, assess. Finally, many of the strategies used by PCMHs, such as team coordination, are not designed around patient safety goals, and thus, cannot reliably be attributed to safety-related motivations.

In conclusion, patient safety issues of greatest concern to clinicians, staff, and patients included communication, EHR interoperability, medication adherence, medication reconciliation, and follow-through on screening and diagnostic tests. Although the perspectives of clinicians and patients were generally aligned about safety issues, patients highlighted trust and communication as most important, highlighting a key focus area to improving patient safety in primary care.

Appendix Table 1.

Semi-Structured Interview Guide

Interview Questions	Probes
What does safe medical care mean to you?	• Is this distinct from how you think about patient safety more generally?
Here is a list of common safety issues that occur in primary care settings like this one. (Provide hand-out with the 4 safety domains)	• Do any of these stand out? • Can you provide an example of something like this happening in a care setting like this one? • Are there other kinds of safety issues in this setting that we have not touched upon?
Are there any practices, processes, or tools at your clinic stand out as being particularly effective in addressing the issues we have just discussed?	• What about safety issues in general?
For safety issues that have NOT yet been addressed, are there practices, processes, or tools at your clinic in place to help catch what you think are compromises to patient safety? If so, are they effective?	• Are these processes applicable to other non-safety related activities or issues?
We are interested in the process of how your practice becomes aware of, or takes actions for safety-related issues. Have you communicated any of the safety issues we discussed to others in your organization? If so, could you describe how?	• Who was involved? • What was/were the action(s) taken? • Would you expect to have sufficient resources (e.g., time, staff) to address this issue? • What is the current situation?
Has anyone in your organization communicated about safety-related issues to you? If so, could you describe how (such as the people, processes, and actions involved?)	• Are there particular processes or activities within your organization for providers to know more about patient safety?
Could you describe the culture in your organization when people share about patient safety concerns?	• Are there any ways to monitor safety-related issues in your practice? • Are there particular processes or activities within your organization to make it easier for people to speak up about safety issues?
What is the patients’ involvement in these safety issues?	• How does the priority of implementing patient safety-related activities compare to other priorities in your organization? For your own work?
Patient-centered medical homes focus on better coordination of care and patient engagement. How do you think the PCMH model is working here?	• What do you think are the main challenges associated with becoming a PCMH?
Now that you’re a PCMH, have you noticed any changes – either for the better or for the worse – in patient safety? Can you provide an example?	• What kinds of information and materials about these [changes in practice] have been made available to you?
With most new initiatives, there are unintended consequences. Are there any patient safety-related issues that either started occurring or became worse after becoming a PCMH?
How do you think teamwork affects the functioning of the PCMH?	• How does this relate to patient safety? • How does leadership, culture, or patient characteristics influence such teamwork?

Appendix Table 2.

Semi-Structured Focus Group Discussion Guide

Discussion Questions	Probes
To start, what words come to mind when you think of safe medical care?
Reflecting on your own experiences, how do your primary doctor and their team communicate with you about your care? Can you give us examples?	• Was there ever a time when a lack of communication caused a problem with your care? • Problems with listening to you well? • Problems with not giving you the right information or not giving it clearly?
Can you describe how your doctor and their team have worked with other medical offices in coordinating your care?	• Lack of working together ever caused problems with care • Not being able to get answers to questions (or the right answer) • Not being told about important test results
Can you describe your past experience in getting a diagnosis?	• Problems getting an appointment at this clinic • Problems getting an answer by phone or electronically • Problems seeing a specialist • Problems getting a procedure or surgery done • Any occasion where you received the wrong diagnosis or were not diagnosed quickly enough
Can you describe the process of how a clinic handled your medication prescriptions?	• Getting the right medication and dose • Being counseled on how to take your medication(s) • Being told about possible serious side effects and what to do • Getting refills • Getting tests that you needed to have done around your medications
How did the process work for you to get other treatments you needed?	• Getting the treatment quickly enough • Getting the right treatment • Having the treatment done correctly
Are there any specific things about this clinic that you think help you get safe care? Can you describe this? (probe PCMH issues specifically)	• Use of electronic health records by the clinic • You having electronic access to your records • Having a case manager, nurse, or other clinic staff help with your care
Who do you go to when you have a problem with your care? This may be related to medication, reaching a physician, getting your medical information, or other issues?	• What the person did to resolve your issue • How quickly they responded • Did you feel like they should have done more to help you? If so, what?
(Ask again to end the session) In one or two sentences, can you tell us once more what safe medical care means to you? That covers the main questions we have for you. At this point, we would like to know if there are any other issues around your safety as a patient you feel would be important for us to discuss?

Appendix Table 3.

Semi-Structured Observation Guide

Observation Categories
General observations and impressions about the clinic, including about:	• High level process map/workflow and how the work system design affects overall process • Organization, culture, team structure • Policies and procedures
Physical Environment: Describe or draw what the physical environment looks like.	• Waiting room • Check in / out areas • Exam rooms • Computer work stations
Tools & Technology (e.g., computers, workstations, phones, paper tools etc.):	• Tool / technology • Location in environment • Who uses (role) • Description (e.g., task(s) used for, content)
Workflow & Processes	• Part 1: Draw each process (workflow) observed (e.g., medicine reconciliation, patient check out). Be sure to label each workflow you draw/differentiate them and if applicable, denote which perspective it is coming from (staff or patient). • Describe the steps for each process (workflow) observed (e.g., medicine reconciliation, patient check out). Be sure to label each workflow you describe and if applicable, denote which perspective it is coming from (staff or patient). Fill in the remaining columns of the table for each process step.
Debrief Questions (asked when the patient encounter was complete)
Questions asked to the staff involved in the patient encounter (excluding physicians)
What is your role and what types of tasks do you typically perform?
What aspects of your work/how you accomplish tasks do you think relate to patient safety?
What are things that you’re aware of that your clinic has done to streamline the medication reconciliation process and make it safer and more reliable?
What are things that you’re aware of that your clinic has done to streamline the patient checkout process and make it safer and more reliable?
Questions asked to the Patient
Overall, how was your experience with the visit today?
Is it clear to you what the purpose of your visit today was? There is no need to share if you do not feel comfortable doing so.
Do you know what your next steps are in your care after this visit (e.g., another appointment, testing, getting medication Rx)? How comfortable/ confident do you feel about the plan for your next steps?
Do you have all of the information you need to move forward with your next steps/in your health care?
What types of information were communicated to you before and during your visit (e.g., “I received instruction on how to take my medications in the room with my doctor”)?
Are there types of information you wish you had or things you wish you knew more about?
Do you use any patient portals (e.g., MyChart, etc.) ? How do you use them?
Do you know the full list of medications you are on?
Did you tell your doctor all of the medications you were on?
Do you know how to take/administer all of your medications?
Do you have any thoughts on how the experiences you had during your clinic visit related to safe medical care?