eMethods. Practice Observation Template
Customize your JAMA Network experience by selecting one or more topics from the list below.
Rubinstein EB, Miller WL, Hudson SV, et al. Cancer Survivorship Care in Advanced Primary Care Practices: A Qualitative Study of Challenges and Opportunities. JAMA Intern Med. 2017;177(12):1726–1732. doi:10.1001/jamainternmed.2017.4747
Why has primary care not widely adopted comprehensive services targeting long-term cancer survivors?
The qualitative data in this mixed methods comparative case study of 12 advanced primary care practices in the United States revealed the absence of a recognized, distinct clinical category of cancer survivorship in primary care; a lack of actionable information to treat this patient population; and current information systems unable to support survivorship care.
Primary care has not adopted comprehensive cancer survivorship services because of the dynamic interplay of 3 factors that must be remedied to change the current situation.
Despite a decade of effort by national stakeholders to bring cancer survivorship to the forefront of primary care, there is little evidence to suggest that primary care has begun to integrate comprehensive services to manage the care of long-term cancer survivors.
To explain why primary care has not begun to integrate comprehensive cancer survivorship services.
Design, Setting, and Participants
Comparative case study of 12 advanced primary care practices in the United States recruited from March 2015 to February 2017. Practices were selected from a national registry of 151 workforce innovators compiled for the Robert Wood Johnson Foundation. Practices were recruited to include diversity in policy context and organizational structure. Researchers conducted 10 to 12 days of ethnographic data collection in each practice, including interviews with practice personnel and patient pathways with cancer survivors. Fieldnotes, transcripts, and practice documents were analyzed within and across cases to identify salient themes.
Main Outcomes and Measures
Description of cancer survivorship care delivery in advanced patient-centered medical homes, including identification of barriers and promotional factors related to that care.
The 12 practices came from multiple states and policy contexts and had a mix of clinicians trained in family or internal medicine. All but 3 were recognized as National Committee on Quality Assurance level 3 patient-centered medical homes. None of the practices provided any type of comprehensive cancer survivorship services. Three interdependent explanatory factors emerged: the absence of a recognized, distinct clinical category of survivorship in primary care; a lack of actionable information to treat this patient population; and current information systems unable to support survivorship care.
Conclusions and Relevance
To increase the potential for primary care transformation efforts to integrate survivorship services into routine care, survivorship must become a recognized clinical category with actionable care plans supported by a functional information system infrastructure.
There are 15.5 million cancer survivors in the United States, a number expected to increase over the next decade.1 Survivors include individuals who have received a cancer diagnosis, are undergoing active treatment, or are in the posttreatment phase.2 The majority of survivors live well beyond their diagnosis, and only one-third continue to seek care from cancer-related specialists 5 years after diagnosis.3 Seventy-five percent of survivors are seen in primary care,3 demonstrating a reliance on primary care to address their needs; however, those needs are not currently being met. A national cross-sectional survey of survivors identified 16 separate domains of unmet needs, with physical needs being most prevalent,4 while a smaller study identified unmet psychosocial and spiritual needs.5 Furthermore, survivors’ rates of preventive and cancer surveillance screening are lower than optimal given increased cancer risk.6
The 2006 Institute of Medicine7(p188) report, From Cancer Patient to Cancer Survivor: Lost in Transition, outlined the 4 key components of high-quality cancer survivorship care: (1) preventive screenings for new cancers, (2) surveillance for recurrence, (3) interventions for late and long-term effects from cancer and its treatments, and (4) care coordination between specialty and primary care. Professional medical groups have begun emphasizing the need for interdisciplinary collaboration to support primary care’s integration of cancer survivorship services,8,9 and the American Cancer Society has released primary care–focused survivorship care guidelines for breast, prostate, colon, and head and neck cancers.10-13
In countries with strong primary care foundations, primary care–based cancer survivorship models have proven effective and acceptable to patients.14-18 In the United States, however, survivorship models rely mainly on specialists’ expertise.19,20 Other proposed models envision integrating primary care clinicians into oncology settings,21-23 but we are unaware of studies analyzing the implementation of cancer survivorship models within community-based primary care practices.
The patient-centered medical home (PCMH) is an American practice redesign model introduced in 2007 to enhance existing primary care practices’ ability to address the emerging complexity of patients’ health needs.24 Core attributes of the PCMH include improved first-contact access to care, more proactive and effective coordination and integration of care, assurance of comprehensive and personalized care, and boosted use of information technology at the point of care.24 The PCMH represents a potential means for primary care to address the comprehensive needs of cancer survivors.25
With few exceptions,23,26-28 research on survivorship care has neglected systems-level issues; however, understanding the primary care context is crucial for developing services that enhance practices’ capacity to deliver population-based care. Despite a decade of effort from national stakeholders to improve the quality of cancer survivorship care, efforts to partner with and support primary care are more recent. It is therefore necessary to assess the confluence of primary care practice redesign and the adoption of comprehensive cancer survivorship services.
This analysis is based on the qualitative data collected as part of a mixed-methods comparative case study of 12 primary care practices throughout the United States. The practices were selected from a national registry of 151 workforce innovators compiled for the Robert Wood Johnson Foundation in 2011 through 2012.29 The registry was developed using snowball sampling by first contacting the authors of 331 articles identified in a literature review of workforce innovations in US primary care since 2000.29 Authors nominated 435 potential practices for initial screening; practice leaders in 227 practices were subsequently interviewed regarding the degree of innovation and sustainability. The existence of cancer survivorship–related activities was not included in the sampling criteria. Deidentified, 2-page practice profiles were constructed from interview notes.
An executive steering committee ranked the profiles to sample for variation in policy context, workforce innovation, and organizational structure. The core research team then clustered highly ranked profiles by state. Lead clinicians or practice managers were invited via email and/or telephone to participate in the study. The first practices were selected in neighboring states (New York and Pennsylvania) for proximity to the New Jersey–based research team for training and data quality control. The next practices were selected for maximum variation in practice type, setting, and geographic region. Recruitment was iterative and proceeded alongside data analysis to search for barriers and facilitators to comprehensive survivorship care. The iterative nature also provided a way to seek out potential disconfirming evidence of emerging findings.
Of the 17 practices responding to our invitation, 12 participated, 2 declined, and 3 were unresponsive after initially expressing interest. Practices provided written consent prior to the researcher’s arrival onsite, and researchers individually consented patients and interviewees throughout each visit. This study was approved by the Rutgers Biomedical and Health Sciences Institutional Review Board.
Over 23 months in 2015 through 2017, 1 of 4 researchers visited each practice for 10 to 12 business days, spending up to 8 h/d conducting ethnographic research and interviews.30 Research trips consisted of 2 visit periods. The first week-long visit included direct observation of practice functions and systems using an observation template (eMethods in the Supplement),30 key informant interviews of practice members,31 and 5 to 10 “patient pathways” of cancer survivors, identified with each practice, who were observed throughout their visit and interviewed about their experiences.30,32,33 Following the Institute of Medicine’s recommendations for who should receive survivorship care,7 survivors were defined as individuals who were no longer in the diagnosis or acute treatment phase (eg, surgery, chemotherapy, radiation therapy).
After the first visit, the researcher debriefed with the qualitative analysis team to determine areas requiring further clarification and to identify practice staff for ethnographic interviews.34 A purposive sampling strategy was used to choose interviewees.35 We sampled for clinicians and administrators in leadership roles (past or current), plus staff who had a multiyear work history and/or represented an instance of workforce innovation (eg, health coach, patient services representative, care coordinator). Researchers conducted 54 digitally recorded ethnographic interviews in total,36 ranging 1 to 3 hours in length (Box 1). Although not all staff were formally interviewed, the researcher’s extended stay at each practice enabled frequent, informal communication with most staff, captured in extensive fieldnotes. Verbatim interview transcripts and practice documents (eg, flow sheets, mission statements, protocols, annual reports) supplemented fieldnotes. The practice manager or lead clinician also completed a Practice Information Form about practice characteristics. The Practice Information Form was originally developed for the Direct Observation of Primary Care Study37,38 and was refined throughout subsequent studies.39
Briefly tell me about the history and evolution of this practice.
Where do new ideas about services come from, and how are they implemented?
Tell me about your current staffing and other unique features of your care team.
What do you see as barriers to care innovation?
Describe how you view your role in cancer survivorship care for your patients.
Where do you refer patients for cancer care? How do you decide where to refer?
What do you know about the survivorship care guidelines recently issued by oncology organizations?
What do you think it would take to put cancer survivorship on your radar?
Data analysis proceeded through a collaborative, interpretive process of “immersion/crystallization.”40 Based on the same inductive reasoning that characterizes grounded theory,41 immersion/crystallization requires iterative cycles of reading the data until salient themes, and variations thereof, begin to emerge. Unlike grounded theory, it does not require the creation of a formal coding structure but rather allows for dynamic interaction among 5 stages of interpretation40 (Figure).
After each visit, debriefing sessions of approximately 1.5 hours were held with the field researcher and the analysis team to review incoming data and identify emerging themes. We used the method by Yin42 of treating each individual case (ie, practice) as a separate study in its own context, which entailed organizing the data from practices and debriefing sessions into case summaries of 5 to 10 pages to highlight distinctive practice features and approaches to cancer survivorship services (Box 2).
The practice’s motivation for change/innovation
The practice’s integration with the larger health care neighborhood
Local options for cancer and other specialty care
The roles and identities of the practice workforce
Nontraditional roles, or traditional roles used in innovative ways
Cancer survivors’ experiences of care in this practice
The practice’s approach to care for cancer survivors
The practice’s comprehensive approach to survivorship care, including favorable conditions and infrastructure, or
The practice’s capacity to implement comprehensive survivorship care (eg, disease registries, care coordination, communication with specialists)
The role of insurance and reimbursement in care
Barriers and/or challenges to implementing change in this practice
Care innovations for other conditions that may translate to cancer survivorship
Two-day analysis retreats with the executive steering committee were held in September 2015 and August 2016 to discuss each practice in detail and begin cross-case comparisons using the case summaries.43 Themes that emerged within and across practices were recorded and consulted throughout subsequent data collection; ongoing debriefing sessions provided opportunities to modify or expand on these themes.44 To validate the accuracy of our findings, practice data, case summaries, and nascent analyses were sent to an external auditor every 3 to 4 months to ensure that findings were minimally affected by preconceptions or bias.45-47
The 12 practices included in this study came from multiple states (Colorado, Illinois, Maine, New York, Pennsylvania, and Washington), medical networks, and policy contexts and had a mix of clinicians trained in family or internal medicine. While not a requirement of the study, all but 3 of the practices were recognized as National Committee for Quality Assurance level 3 PCMHs (Table).
None of these practices had any comprehensive cancer survivorship services. Instead, survivors received care equivalent to that of nonsurvivors. We identified 3 themes across all 12 practices that speak to potential barriers of implementing comprehensive survivorship care: (1) clinicians do not treat survivors as a distinct population or clinical category; (2) clinicians receive limited information or follow-up guidance on patients’ cancer care; and (3) current information systems are insufficient to support survivorship care.
Clinicians had difficulty articulating what made cancer survivorship a meaningfully differentiated clinical category. There was confusion over what survivorship care entailed beyond monitoring for recurrence, as well as how to measure the care provided. As 1 physician in practice 11 (P11) said, “We’re always going to be attendant to things that are obviously important but also measurable. So what would you measure with cancer? Is it their holistic care, their emotional stability?”
Some clinicians believed that their care was comprehensive enough to address the needs of all patients, without special consideration for cancer survivors. The lead physicians in P1 and P4 identified cancer as a chronic medical condition to be monitored regularly like other chronic conditions, with P1’s lead physician commenting, “There are more patients who live with cancer or who live with having survived cancer, and it just becomes part of their chronic medical illness.” The lead physician in P4 echoed this sentiment, saying, “I certainly don’t think of [these patients] as cancer survivors. I think of them as people living with a history that’s not terribly unlike the history of diabetes or any other chronic disease.”
In most practices there seemed to be a different clinical norm for cancer compared with other conditions, a kind of “cancer exceptionalism.” On diagnosing cancer, clinicians referred their patients to oncology and soon became disengaged; the cancer diagnosis superseded all other medical conditions, effectively rendering oncology responsible for the totality of patients’ care. As a P8 physician said, “The patients pop in, they have cancer, we see them, we take care of their problem, but really, it has been—almost like the psychiatrists have carved out mental health, I think.” Primary care clinicians lost touch with these patients as they “disappeared” into oncology, and survivorship remained a concept with little practical meaning in the primary care setting.
Some practices received cancer-related information about their patients, but it was either inadequate or not actionable. Practice 2 was located in one of the most highly cancer-resourced areas in the country, which included 4 National Cancer Institute–designated cancer centers within a 30-mile radius. The practice was networked with 2 of these centers, which facilitated obtaining patients’ oncology records. Still, these records did not provide enough information for primary care generalists to take appropriate actions. As a P2 physician said, “I want to make sure I understand what treatment they had, but it kind of happens in a black box…, so I feel a little intimidated by it, to be honest with you.”
In addition to their unfamiliarity with many cancer treatments, primary care clinicians described limited guidance for long-term follow-up care. One P3 physician noted how quickly oncologists’ discharge summaries became outdated, as they provided guidelines for the “next immediate step” but failed to address potential future concerns. Another P3 physician said she learned about patients’ cancer histories through “active questioning” at annual examinations—but only if there were not too many recent problems to address. By then, the cancer might have occurred so long ago that communicating with the oncologist was no longer possible. One of her patients had reported receiving a hysterectomy in 1985 to treat her cervical cancer. “What are you supposed to be doing for your follow-up of that?” the physician asked. “Realistically, it’s that many years ago. Maybe it doesn’t play into it at all, but I don’t have any guidance for that, really, so you kind of make it up on your own.”
All 12 practices had well-established electronic health records (EHR) systems; however, these EHR systems were inadequate for implementing population-level interventions in cancer survivorship. Clinicians mentioned an inability to identify patients with a history of cancer, which led to a further inability to create a registry and track survivors as they could track other patient populations.
Patients were responsible for reporting their cancer histories to their clinicians, and medical records were sometimes lost as patients changed clinicians over the years. Even when a patient’s cancer history made it into the EHR, it was not always easy to find. Practice 5’s nurse practitioner noted that EHR problem lists were not “user-friendly,” which meant that she had to search through multiple screens to find any mention of cancer. “In the olden days, when you had a paper chart and you opened it, the problem list … was on the left side, and it was handwritten,” she said. “And it might be 2 or 3 pages, but it was there at a glance. With the electronic medical record, sometimes it’s a needle in a haystack.”
In addition to tracking survivors, there was a need to monitor changing information in oncology. A P6 physician noted, “I think we [in primary care] would like to take on that role as custodian of the survivorship plan, knowing that we would always have access to the oncologist. But I think the outstanding question is how do we make sure those plans are up-to-date and front-and-center and not lost in the busyness of modern medical records?”
Other clinicians observed that an inability to track survivors through the EHR made the extent of their survivor population unknown. Some EHRs lacked any specific acknowledgment of cancer histories, and tracking survivors meant manually creating an ancillary database outside the EHR. According to 1 P5 physician, an important implication of not tracking cancer survivors was the difficulty in identifying populations at risk for treatment sequelae or secondary cancers: “I think when you read books about cancer, and you read that there are all these secondary cancers, and you sit here and think, ‘Gosh, I wonder if any of my patients could get secondary cancers,’ there’s no way to look it up.”
Despite the push from national organizations to enhance cancer survivorship care capacity in primary care, findings from this study suggest that cancer survivorship care does not integrate easily into advanced primary care. At this time, there seems to be no recognized receptor site for cancer survivorship services in primary care.
Survivorship remains a diffuse concept with multiple meanings that may complicate clinical care.48,49 Survivorship lacks clear codified diagnostic criteria, care protocols, and expectations, without which clinicians had no sense of what survivorship care looked like. In the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), the closest diagnostic code for cancer survivorship is Z85, an aftercare code indicating personal history of malignant neoplasm that is not directly billable. Codifying survivorship as a distinct clinical category that belongs on problem lists with payment-linked (fee, value based, or capitated) care services is a critical first step toward bringing comprehensive cancer survivorship services to primary care.
As noted in the Results (theme 1), primary care clinicians let go of their cancer patients more easily than other patients. Their comments about oncology referrals were different than those about other referrals, in which conversations addressed the boundaries and expectations of shared vs consultative care. With cancer, there was a tacit willingness to concede all care to oncology. We refer to this distinction as cancer exceptionalism because the diagnosis of cancer seemed to override all other problems. A diagnostic code for survivorship may be a first step toward changing cancer’s exceptional state.
Once primary care recognizes cancer survivorship as a distinct clinical category, there need to be easily accessible, actionable care strategies.23 Primary care clinicians have different needs than oncologists in providing survivorship care,27 yet those needs remain unacknowledged in current survivorship care planning discussions. A British study on the relevance of survivorship care plans to general practitioners suggested incommensurability between oncology-driven guidelines and the realities of primary care.50 A disease- (and cancer site–) specific model, while suitable for specialists, contains too much detail for a generalist.51 Most primary care patients, including cancer survivors, have multiple comorbidities, thereby requiring a comprehensive approach to care.52 Highlighting specific risk information related to secondary cancers and treatment sequelae, along with recommended surveillance for recurrence, may prove more useful.
Without a diagnostic code for survivorship on the patient’s problem list or customizable protocols to remind clinicians of necessary follow-up, the EHR remained more hindrance than resource. As EHR clinical decision support tools become increasingly powerful drivers of clinical care,53 the lack of developed supports for survivorship means that survivorship cannot compete for attention with chronic diseases that already have robust technologies and quality measures.54,55
This study is not without limitations, specifically in its small sample size. Furthermore, while ethnographic interviews demonstrated a gap in clinicians’ knowledge about the Institute of Medicine’s 4 key components of survivorship care,7 this study did not directly assess this gap, nor did it assess clinicians’ baseline knowledge of cancer survivorship. Regardless, findings about the lack of comprehensive services were consistent across practices, all but 3 of which were well-established level 3 PCMHs from diverse policy and organizational contexts. Further research is necessary to determine whether these results apply to other contexts.
Comprehensive cancer survivorship care is not happening in advanced PCMHs. The absence of this care is from the lack of concrete solutions to needs that precede the development and implementation of models of care. There is no discrete clinical category of cancer survivorship for primary care. Even with such a category, primary care clinicians rarely receive actionable information to care for this patient population. Finally, actionable information for individual patients does not translate to population health because of the lack of adequate information systems. By correcting these 3 deficiencies, comprehensive cancer survivorship services could more easily integrate into current primary care transformation efforts.
Corresponding Author: Ellen B. Rubinstein, PhD, Department of Family Medicine, University of Michigan, 1018 Fuller Street, Ann Arbor, MI 48104 (email@example.com).
Accepted for Publication: May 22, 2017.
Published Online: September 25, 2017. doi:10.1001/jamainternmed.2017.4747
Author Contributions: Drs Rubinstein and Crabtree had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Miller, Hudson, Howard, Crabtree.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Rubinstein, Miller, Hudson, Howard, O'Malley, Tsui, Crabtree.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Bator.
Obtained funding: Crabtree.
Administrative, technical, or material support: O’Malley, Lee, Bator.
Supervision: Miller, Hudson, Crabtree.
Conflict of Interest Disclosures: None reported.
Funding/Support: Funding for this research was provided by National Cancer Institute grant R01 CA176545, “PCMH Implementation Strategies: Implications for Cancer Survivor Care” (primary investigator: Dr Crabtree).
Role of the Funder/Sponsor: The National Cancer Institute had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We thank the members of the executive steering committee (Kurt Stange, MD, PhD, Case Western Reserve University; Kevin Oeffinger, MD, Duke Cancer Institute; Paul Nutting, MD, retired; Carlos Jaen, MD, PhD, University of Texas Health Science Center; and Larissa Nekhlyudov, MD, Harvard Medical School) for their input on our study design and sampling framework, as well as the external auditors (Kris Gowen, PhD, and Elisabeth Hicks, MA, Oregon Health and Sciences University) for interpretation of data. All individuals were compensated for their time.
Create a personal account or sign in to: