Meeting high-risk patient pain care needs through intensive primary care: a secondary analysis


  • We leveraged existing data from a five site demonstration project to better understand primary care based pain care for high-risk patients.

  • The number of PIM participants involved in the study is substantial.

  • Our methods allowed us to explain both pain prevalence as well as stakeholder perceptions.

  • Interview questions did not directly target pain or pain management, however, impacts on pain care was a central subject of many of the interviews.

  • The approach we used allowed us to understand themes separately by stakeholder group.


Chronic pain impacts more than 100 million adults in the USA1 and has deleterious consequences across many aspects of patients’ lives.2–4 Many of these patients have complex healthcare needs, including both mental and physical health disorders that put them at high risk for near-term hospitalisation or death, a status that can be identified using electronic medical record tools such as the Veterans Affairs (VA) Care Assessment Needs Score.3 5–8 Yet, the intersection of chronic pain and high risk of hospitalisation has seldom been specifically addressed, either in terms of its frequency or its intervention impacts, in studies of interventions for improving outcomes among high-risk patients.

High-risk patients are predominantly cared for in general primary care (PC) settings.9 In turn, appropriate care for these patients consumes substantial PC resources.10 11 Chronic pain care is known to be challenging for PC settings,2 even within patient-centred medical home models such as VA’s Patient Aligned Care Teams (PACT).1 10 12 Intensive management (IM) teams have been tested as an approach to supplement and support PC for high-risk patients,13 14 yet these studies have rarely focused on chronic pain and its management. To address how pain and its management may impact IM interventions on high-risk patients, we conducted secondary qualitative and quantitative analyses on data from an evaluation of IM teams in PC conducted in 2014–2018.15 PACT Intensive Management, or PIM, was a multisite VA quality improvement initiative carried out in five VA healthcare systems. Prior qualitative studies15–17 found that PIM patients frequently reported pain care needs and identified difficulties with care coordination.18 However, neither the PIM evaluation, nor the IM intervention, had a specific focus on pain or pain care.

Our study goal was to assess the importance of developing a focus on pain within studies aimed at improving outcomes among high-risk patients. Our objectives for the study were to (1) determine prevalence of pain among high-risk Veterans and (2) understand what patients, PC providers and IM teams perceived as the impacts of IM on pain or pain management.


Overview: To assess pain frequency among high-risk Veterans, we carried out secondary analysis of PIM evaluation data from 2014 to 2018 on high-risk patients from participating VA integrated healthcare systems. To assess possible IM impacts on pain, we used secondary analysis of data from PIM evaluation interviews of patients and PC providers exposed to IM, and with IM team members.

Study design

PIM used national data from VA’s weekly assessment of Veteran Care Assessment Need (CAN) scores11 15 19 across PC and women’s health clinic patients to identify those (a) within the top 10% risk category by CAN score and (b) who had a hospitalisation or emergency department use during the prior 6 months (3794 patients). 1904 of these patients were randomly selected as eligible for an invitation to participate in IM, along with 201 patients referred to IM by their PC providers. The resulting 2105 patients constituted the IM intervention group; 1890 were assigned to usual care.

IM teams at five geographically diverse healthcare systems aimed to provide coordinated care to high-risk patients as a supplement to general PC, based on PIM study intervention guidance and materials.16 Teams typically included an internist, nurse, social worker and mental health provider.13 15–17 20 21 The teams performed comprehensive assessments, preventative home visits, transitional care management, medication management, care coordination, health coaching, patient and caregiver education, case management for social needs, and advance care planning.16 Participating systems had substantial autonomy in designing the details of their IM approaches.

Assessment of pain severity and prevalence

To assess pain severity, we used VA electronic medical record data to access the pain scores collected from all Veterans in VA PC as part of routine vital signs from 2014 to 2018. Pain scores are assessed on a 0–10 scale, with scores 4–10 indicating moderate to severe pain.22 We assessed the median and mean pain scores for the intervention and usual care patients for the year prior to random assignment and 6–18 months after random assignment. We measured prevalence of chronic pain conditions for intervention and usual care patients for the year preceding random assignment to IM and over the patient’s lifetime (ie, the earliest data available in VA electronic medical record data up until the date of IM assignment) through ICD-9 (International Classification of Diseases) and ICD-10 codes.

Qualitative assessment of IM impacts

We identified pain mentions within 104 PIM interview transcripts focused on organisational participants (eg, providers, patients) exposed to the IM intervention. Interviews occurred between March 2015 and October 2017 and lasted between 20 and 30 min.

PIM evaluators purposively sampled and invited 78 patients who had participated in IM for interviews on IM care and its impacts, of whom 51 completed interviews. Evaluators also approached 35 IM team members, 29 of whom completed interviews, and 34 PC providers (physicians, nurse practitioners or physician assistants), 24 of whom completed interviews.

To conduct the interviews, PIM qualitative investigators used semistructured interview guides with questions that were developed iteratively by a group of experts in mental health, PC and qualitative methods. Interviews were audio recorded, professionally transcribed and deidentified.

The PC provider and IM team guide covered the characteristics and process of IM as perceived by interviewees. The patient guide covered patient experiences with and perceptions of IM, including overall satisfaction with the IM programme. Questions from these guides are documented in previously published work.17 23

To identify the interview transcript excerpts that mentioned pain, we used the qualitative analytic software ATLAS.ti 724 autocode function to analyse all 104 PIM evaluation interviews. We defined pain mentions as any reference to the word ‘pain’ or its synonyms, including ‘hurt’ and ‘pain medications,’ such as ‘narcotics’ and ‘opioids.’ We then extracted the full paragraphs surrounding the pain mention. A research assistant trained in qualitative methods (JRH) then manually reviewed the flagged transcripts and compared pain mentions with the Atlas autocode results, and then resolved inconsistencies. The final analytical data set, thus, included all identified pain-related content from the transcripts. Authors of this paper (KG, MM and JRH) conducted content analysis by independently reviewing paragraphs with pain mentions, then sought consensus on themes through four rounds of discussions, comparing the themes by organisational participant type using constant comparison.25

Patient and public involvement



We identified 3723/3858 high-risk Veterans (96.5%) as having accessible electronic medical records and at least one pain score during the specified timeframe. Pain prevalence was common among these patients. Among the 3858 high-risk Veterans, nearly 70% (2593/3723) experienced moderate to severe pain (a pain score of 4–10),22 with an average pain score of 2.8 at baseline. Most high-risk Veterans had a chronic pain condition in their lifetime; 86% had been diagnosed with arthritis, 77% with chronic neck or back pain.

Among the 104 PIM interviews, there were a total of 89 pain mentions, distributed among the 40 interviews that mentioned pain at least once. Among the pain mentions, 42% (37/89) were contributed by 18 patients, 24% (21/89) by 11 PC providers and 35% (31/89) by 11 IM team members. The 11 IM team members included 2 physicians, 3 nurses, 3 psychologists, 2 social workers and 1 medical support assistant (see table 1).

Table 1

Participant demographic characteristics

Our analyses of the 40 interviews that included a pain mention identified 7 themes within 3 general domains that categorised the respondent’s point of view as representing the perceptions of patients, PC providers or IM team members regarding the impacts of the IM programme on pain care. Patient-identified themes were that IM improved communication and responsiveness to pain. PC provider-identified themes were that IM improved workload and access to expertise. PIM team member-identified themes were that IM improved pain care coordination, facilitated non-opioid pain management options, and mitigated provider compassion fatigue. No negative IM impacts on pain care were mentioned.

Patient perspectives on the impacts of IM on their pain care experiences

Improved communication

The patient commented on IM team support for patient–provider communication about pain care. Patients also mentioned IM team support which improved coordination and communication about pain among their multiple providers, noting that ‘they’re [IM providers] really good about communicating with them [other providers/specialists], too, which is something I’ve always wished the VA would do is communicate with each other as far as doctors go.’ Patients identified the improved communication as helping to reduce their frustration with pain care, saying that working with IM providers‘made a difference in the fact that I have somebody I can talk to without getting all frustrated’ by discussing issues that impacted their chronic pain and overall health. Patients reflected on feeling unable to express pain to their PC providers because the pain was ‘not understood, between the PACT and my primary care…I’m asked to explain things but as I get into the moment…I get shut down. So if I can’t express my feelings, how I feel in the moment…I’m not really expressing my true feelings and that adds to the aggravation.’ Patients also indicated that the IM team did a better job encouraging patients to express their pain-related experiences. Patients felt IM facilitated sensitive communication around pain care, since pain is a ‘deeply personal’and ‘subjective’ experience. One patient described having ‘issues communicating just how much pain I’m in and getting treatment and help from the VA [prior to IM].’One patient summarised that their IM team was ‘making a difference literally in my health, my leg, my pain level. I want to stay with that.’

Responsiveness to real-time pain situations

The patient additionally identified the importance of both the IM team and PC providers for addressing practical, time-sensitive pain care-related issues. Help with these issues included coordinating quicker access to pain care. One patient reported challenges with accessing his/her PC provider in response to a severe pain episode. His PC provider took him to the scheduler and said, ‘If he calls you, you set an appointment. I don’t care if you have to double-book me, triple-book me or whatever. If he calls you, you set an appointment within a week.’ Six weeks later, the patient called the office scheduler and ‘asked her [about] an appointment, I was in severe pain.’ Despite the PC provider’s best intentions, the scheduler said, ‘The next available appointment is three months out.’ The patient had more confidence in the ability of IM to meet his severe pain needs, reporting being‘thankful that I don’t have to go back to that [PC] clinic any more right now [because I am in IM].’ Another patient commented that pain is a symptom that requires real-time responsiveness and expressed dissatisfaction prior to IM enrolment: the healthcare team ‘isn’t compassionate to the fact that you’re in pain and you’ve been getting the runaround [and have] called three or four times [and] didn’t get a hold of anybody.’

Patients also commented on the additional time and work from the IM team to provide a timely and satisfactory response to pain crises. One patient, during a visit with her PIM nurse, described that ‘it was so painful that when I was talking…I would have to stop for a moment and just emotionally get myself back together again before I could go further on how I was explaining how this pain was and it was so severe.’The IM team then responded by ‘really working with me in trying to find out what is actually causing this pain… [and now it is] just minute for minute with me by the knowledge and the contact, they are able to come up with the results that answers quickly.’One patient clearly detailed how differently his pain was acknowledged and validated within IM compared with usual PC:

‘And I was hurting, so, finally, when they got to me I told them,‘I’m not sure what I’ve done but the thorax part of my back,’ I said, ‘I sneezed and it’s hurting.’ And they treated me as if it was no big deal. I was set to one side and finally the doc came up to me and it was like I was wasting his time…I wanted to tell him so bad, you know, your mannerism almost got—if I was not who I am, I would’ve punched you out… And when I left from there and came home, I felt so sorry of how Vets are being treated because of that one factor right there. But since I’ve gotten with these people here [in the IM program] I don’t feel that. I don’t see that. I’ve never seen any more of what I’ve just described. What I described was long before I got onto this [IM] program. This program is like a 180 from what I’d been seeing.’

PC provider perceptions of IM contributions to pain management


PC provider interviewees who mentioned pain and pain care commented on the helpfulness of IM teams for reducing the PC workload around meeting the needs of high-risk patients both in general and specifically for pain: ’…Veterans that have needs that the [PC] team cannot put their attention on and help them, then our IM team is available to assist the [PC] team.’ Specifically, IM allowed longer clinic visits to accommodate needs of high-risk patients since ‘generally, with [PC] … it’s a very quick turnaround visit to get them in and to get them out.’ On the subject of pain, PC providers noted that they often found themselves feeling that they ‘have a Veteran that just needs more time’. Other PC providers noted that IM reduced the PC team workload associated with providing pain care in the context of mental illness and substance use. This benefit of IM, however, did not always last once such a patient returned to PC. One PC provider gave the example of a patient with pain, mental illness and substance abuse who had been doing very well in the PIM programme for 6 months to a year. However, when the patient was discharged from IM, the ‘problems that had happened before resurfaced, because he just really needed a lot of handholding because of mental illness and substance abuse and other chronic pain… once he was discharged back to me… the workload returned to the way it had been prior.’

Access to pain expertise

PC providers noted that in addition to providing help with workloads, IM teams provided access to expertise on subject matters that improved outcomes for their high-risk patients, including pain. Multiple PC providers indicated lack of confidence around pain management, particularly around understanding and dealing with the side effects, monitoring and tapering issues among patients on opioid medications and described the benefits of IM in this regard. One PC provider described how IM helped strategise creative processes for taking opioids with a coprescribed stool softener and ultimately reduced emergency room (ER) visits. The PC provider said, ‘This [safe and effective opioid pain management] is where I need the help,’and an IM team member helped to ensure that the high-risk patient got ‘the pain meds throughout the day and…docusate and senna throughout the day, so he hasn’t been in the ER with pain complaints of his back, and hasn’t been in the ER with complaint of constipation because he’s getting those meds now.’ Multiple PC providers also described benefiting from access to expertise on high-risk patient care from IM teams. One PC provider described a patient with early stages of dementia as well as back and abdominal pain who was always going to the ER, however,’since they’ve [the patient and his spouse] come on board [with] the [IM team], I think he’s only been into the ER once.’ Other interviewees gave examples of the positive impact of IM support on provider comfort with navigating pain care.

IM team perceptions of their roles in pain management

Pain care coordination

IM team interviewees regarded support for pain management among high-risk patients as an important role, including support for pain care coordination. IM team members identified pain as one of the reasons PC providers want to refer patients for IM and identified management of chronic pain as an important area on which to focus. For example,‘helping them [IM patients] feel like they can navigate and coordinate a very complex system more effectively’ is an important element in IM. Additionally, ‘improving their health by maybe better assurance or other interventions that are going to be helpful for them…[such as for] chronic pain management’can be an important contribution of IM.

Facilitation of non-opioid pain management options

IM team members identified facilitation of non-opioid pain management strategies as an important contribution. Team members referenced the dedicated time they were able to spend with their patients on this issue through the IM programme, as well as their relevant expertise, as allowing for increased trust and relationship-building with PIM patients. One patient joined IM entirely disengaged from his healthcare,‘But after meeting with us [the IM team] and sitting down and seeing that we were going to do things differently, we were going to actually listen to him, he trusted our recommendations as far as like the pain medicines go.’This trusting patient–provider relationship within IM helped patients engage in their healthcare, and this, in turn, encouraged patients to try non-opioid pain management strategies. IM team members cited the ‘rapport building was very important’ and that rapport allowed patients to develop‘trust’ in their PIM providers, enabling patients to explore non-opioid pain management options. For example, when the IM team recommended non-opioid therapy for one patient in pain, the patient‘trusted our recommendation and…decided to do hypnosis for pain and…CBT for pain.’

Mitigation of provider compassion fatigue

IM team members who mentioned pain also identified mitigation of PC provider compassion fatigue as an important IM benefit. IM team members described that ‘people get worn out on…intensive patients’ and may require emotional backup when compassion fatigue is high. One IM team member described a PC provider’s conversation with a PC nurse during which the nurse said, ‘[I] just can’t really empathize with this person anymore.’ The IM team member noted that this feeling ‘sometimes affects their [PC providers’] ability to really help that patient.’Team member comments implied that IM can offer an extra level of support, or a safety valve, which may be critical in mitigating this kind of compassion fatigue, particularly for patients with chronic pain.


We postulated that pain would be a frequent and important symptom among high-risk patients and sought to identify whether IM-enrolled patients, their PC providers, and IM teams would spontaneously identify IM impacts on pain care. We found that moderate to severe pain is common among high-risk patients, and of such concern to patients, PC providers and IM teams that randomly selected interviewees from each of these groups frequently mentioned pain and/or pain care when commenting on IM, even though interviewers did not ask about these topics. Based on interviews, we found that PC providers struggled with caring for high-risk patients with chronic pain due to lack of comfort with chronic pain management, whereas IM teams were able to provide faster access and spend more time with patients with chronic pain. We also identified seven themes that can be used to guide future care for high-risk patients.

PC-based IM for populations of patients identified through electronic data as being at high risk of hospitalisation aims to preventively intensify care, thus avoiding deteriorations in patient health that could result in the need for hospitalisation. Results from IM programme evaluations, however, show mixed results.13 17 The IM programme on which the analyses in this manuscript are based,15 found that outcomes among randomly invited and non-invited patients showed no significant reduction in Veterans Health Administration (VHA) emergency department visits, hospitalisations or overall costs.26 Results did, however, show enhanced care coordination and improved patient experiences at no greater cost.13 18 20 Our qualitative results provide further context for these quantitative results, showing that patients with pain experienced IM as improving pain-related care coordination and satisfaction with care.

Our findings gain importance through the known link between chronic pain management and healthcare provider burn-out10 12 27–29 as well as noted PC provider discomfort with elements of pain care provision.30–33 One of two PC provider themes focused on IM team ability to reduce the workload related to appropriate pain management among high-risk patients, given complex accompanying physical and mental conditions. PC providers thought that IM enabled patients with pain to benefit from longer visits, smaller panel sizes, and easier access to consultation with both IM team and non-IM team specialists than could be provided within routine PC. Interestingly, access to expertise could include, but was not limited to, access to pain specialists. PC providers also credited IM team members with time and expertise to carry out creative problem solving around a patient’s pain care needs, showing that meeting these needs went beyond knowing the right therapies to prescribe or recommend.

IM team member interviews help us to understand how the IM teams attempted to improve pain care through the themes of coordinating care, facilitating non-opioid pain management options and mitigating provider compassion fatigue. IM team members mentioned how the structure of IM allowed them to facilitate non-opioid and/or non-pharmacological pain management approaches; transitioning patients to non-opioids is known to require more provider time and engagement than does refilling an opioid prescription.34 IM teams identified the need to use visit time to build rapport and trust as a key element in engaging patients to try something new, such as cognitive behavioural therapy or acupuncture. Similarly to the PC provider interviews, IM team member interviews related PC access to IM as potentially decreasing PC provider compassion fatigue.27

The two patient interview themes (communication and real-time responsiveness) triangulated well with those derived from PC providers and IM team members. Patient-reported improvements in pain-related communication through IM included comments potentially related to care coordination, including facilitation of communication between patients and PC providers and between multiple providers about the patients. Patient interviews also highlighted the value of improved communication around their pain-related issues as a positive aspect of IM, including for relationship-building, for validation of the pain experience, for discussion of pain impacts on quality of life and for developing mutually agreed on pain care plans. Improved communication through IM team pain care likely also relate to IM team themes of relationship-building around pain, of introducing new approaches to patients’ pain care, and of mitigating PC provider compassion fatigue. Both the patient-generated theme of communication improvement and of improved real-time responsiveness to pain-related concerns through IM are potentially related to PC provider comments on workload shift to IM.

Our study has limitations. First, our conclusions are based on secondary qualitative analysis; interview questions did not directly target pain or pain management. Second, while interviewees were randomly selected from each of the three target groups (patients, PC providers and IM teams), and while the full interview sample of 104 was substantial, our results are preliminary. Of the 104 interviews, only 40 mentioned pain. The number of interviewees within each of our target groups is, therefore, small and cannot fully represent each group’s perceptions. Our findings, however, reflect synergy across the groups’ themes, strengthening our conclusions. Finally, our study is observational and cannot identify causation between IM and the themes we identified.


In conclusion, our findings show the importance of pain and pain care among high-risk patients both quantitatively and qualitatively. Moderate to severe pain afflicted 70% of a diverse population of high-risk patients and generated 89 spontaneous mentions of pain-related concerns across patients, PC providers and IM teams in interviews. Our findings suggest that inclusion of a focus on pain and pain-related care may be critically important in future efforts to improve high-risk patient care and outcomes.

Data availability statement

No data are available. The data are not publicly available due to their containing information that could compromise the privacy of research participants.

Ethics statements

Patient consent for publication

Ethics approval

PIM was conducted as a non-research operations35 activity designed to improve programme operations, under the authority of the VA Office of Primary Care.36 Because this was designated as a quality improvement project, there was no formal informed consent process. This secondary data analysis is part of the quality improvement project and was selected as a focus with the approval of the VA Office of Primary Care. The interviews, however, do include an introduction regarding the purpose of the interview and assurance of confidentiality. Permission to audio record the interviews was requested.

This post was originally published on