Abstract
Objectives
Over two-thirds of older individuals live with multiple chronic conditions, yet chronic diseases are often studied in silos. Taking a lifespan approach to understanding the development of multiple chronic conditions in the older population helps to further elucidate opportunities for targeted interventions that address the complexities of multiple chronic conditions.
Methods
Semi-structured interviews were conducted with 38 older adults (age 64+) diagnosed with at least two chronic health conditions. Content analysis was used to build understanding of how older adults discuss the timing of diagnoses and subsequent self-management of multiple chronic conditions.
Results
Findings highlight the complex process by which illnesses unfold in the context of individuals’ lives and the subsequent engagement and/or disengagement in self-management behaviors. Two primary themes were evident regarding timing of illnesses: illnesses were experienced within the context of social life events and/or health events, and illnesses were not predominantly seen as connected to one another by patients. Self-management behaviors were described in response to onset of each illness.
Discussion
Findings provide insight into how older adults understand their experience of multiple chronic conditions and change in self-management behaviors over time. In order for practitioners to ignite behavioral changes, a person’s history and life experiences must be considered.
Poor health outcomes and high health care costs for individuals living with multiple chronic conditions (MCCs) are well documented.1,2 Approximately one in four American adults and two-thirds of Medicare beneficiaries have MCCs (i.e. two or more health conditions slow in progression, long in duration, and devoid of spontaneous recovery); 14% have six or more health issues.3–5 Evidence-based recommendations for care and self-management of MCCs are complex due to the multitude of factors involved (i.e. recommended health behaviors, personality dispositions, past response to health problems 6 ). As a result, many older people are uncertain of how to manage and adapt to their multiple illnesses.
Recently, the United States Department of Health and Human Services (HHS) and the National Institutes of Health (NIH) established an initiative specifically focused on the prevention and management of MCCs. 7 However, current literature and practice guidelines do not address how the timing of diagnoses and use of self-management strategies unfold across the lifespan of an individual. For example, it is likely that the diagnosis of heart disease early in life, followed by a diagnosis of diabetes, and then depression has a very different impact on an individual than if the person is first diagnosed with depression, then diabetes, and then heart disease. How individuals acclimatize to new illnesses over time, as well as the sequence of those illnesses, may alter health outcomes. Furthermore, contextual factors in an individual’s life may impact an individual’s response to a particular illness or combination of illnesses more so than others. Taking a lifespan approach to studying the development of illnesses in the older population can advance our current understanding of how to support individuals with MCCs as they age. This study utilizes in-depth, semi-structured interview data to descriptively examine how older adults narrate their development of MCCs and resultant behavior changes after receiving each diagnosis.
A lifespan approach to understanding MCCs
Over the past decade, considerable research has been conducted to enhance our understanding of the impact of chronic illnesses on health outcomes. However, much of the work has focused on the study of individual illnesses,8,9 as counts, 10 or as dyads and triads of illnesses.11–13 Yet, older adults experience varied combinations of upwards of six or more illnesses that develop over a lifetime.14–16
A lifespan developmental perspective posits that social, behavioral, and biomedical processes interact to shape responses to illnesses over time and that these responses play a critical role in shaping future health. 17 Taking a lifespan approach to understanding MCC progression puts the person’s contextual life circumstances (i.e. family dynamics, financial concerns, explanatory models of disease, medical knowledge) and prior illness history at the center of understanding illness impact. It suggests the need to guide care at the individual-level and to consider initiation and maintenance of self-management behaviors based on how illnesses unfold and interact within the context of an individual’s life. The lifespan approach also suggests the importance of understanding how certain sets of diagnoses can have compounded effects or no effect on a person’s life. However, limited work has drawn on a person-centered approach to studying MCCs. 18 Roberto and colleagues 19 used a life-course perspective to understand the role of MCCs in the lives of older adults to see how their life history of illness impacted their beliefs about health. Sells and colleagues 20 further addressed individuals’ understanding of their illness onset over time. However, additional research is needed to further advance our understanding of how older adults perceive the unfolding of MCCs over time and their subsequent behavioral responses to each illness within the context of their other illnesses.
Self-management of MCCs in older adults
Recent work has drawn attention to the varying levels of self-management required by individuals to cope with chronic illnesses. 7 Self-management strategies include behaviors such as diet, exercise, and monitoring addictive behaviors. For example, modified diets have shown benefits in prevention of chronic diseases 21 and are often recommended for self-care for illnesses such as hypertension, heart disease, and/or diabetes. Physical activity can reduce the risk of developing chronic conditions such as diabetes 22 and cardiovascular disease 23 and has an ameliorative effect on depression. 24 Smoking and excessive alcohol use carry further implications for treatment and care of illnesses such as pulmonary or heart disease. 25
However, global guidelines for implementing self-management behaviors in response to chronic illness are complex and encourage individuals to engage in a series of behaviors for each illness they have.6,26,27 Such a strategy can overwhelm the patient and often fails to consider how different guidelines may overlap, interact, or contradict one another in the context of care for multiple illnesses, 28 or whether patients are able to or are invested in making behavioral changes in their lives. Little work has examined how self-management behaviors unfold over time within the context of MCCs. For example, it might be that onset of a certain illness in a particular order triggers a behavioral response more so than other illnesses. Understanding how older adults describe their uptake of self-management behaviors over their life course of illness development may prove critical in informing concrete, practical care recommendations for older adults with MCCs and streamline guidelines for illness combinations that may arise.
Taking a qualitative approach to understanding experiences of MCCs in older adults
In order to build a ground-up perspective and learn from individual-level experiences, qualitative methodologies are essential. Relying on the voices of participants through open-ended questions can provide valuable insight without introducing researcher bias. 29 Past qualitative work has examined the experience of older adults living with MCCs while managing the demands of specific chronic health conditions. 28 Additional work has examined patients’ illness representations in the multimorbidity context 30 and the role of social support in bolstering resilience and adaptation when considering the cascading effect of acquiring MCCs. 20 Still others have explored patient perceived barriers to self-management of hypertension 31 and the cognitive representations and self-management of diabetes and depression 32 . Yet, no study to date has specifically explored how older adults narrate their experience of MCCs unfolding in their lives and how they subsequently describe adjusting self-management behaviors over time. Participant narratives may elucidate and identify adaptive responses to living with MCCs.
Current study
The purpose of this study was to use in-depth, semi-structured interview data with older individuals (64+) with two or more chronic health conditions to expand our understanding of how older adults narrate their development of chronic conditions and subsequent change or maintenance of self-management strategies (i.e. nutritional habits, exercise regimen, medication use, and addictive behaviors). Participant transcripts were then analyzed used Content Analysis to identify key themes in narratives.
33
Like Roberto and colleagues,
19
we take a lifespan approach to understanding the progression of MCCs in older individuals, but focus on narratives of onset and self-management strategies. Our research goals were to:
Develop an understanding of how older adults narrate their experiences with the progression of MCCs over time; and Learn how older adults describe changes in their self-management behaviors over time in the context of the development of MCCs.
Methods
Sample
Data were obtained from 38 adults aged 64 and over who (1) identified as having two or more chronic health conditions, (2) were receiving medical care at The New Jersey Institute for Successful Aging (NJISA), a geriatric clinical care center in Stratford, New Jersey, and (3) were capable of completing an hour-long interview in English. See Table 1 for demographic characteristics of the sample. Participants varied in their type and number of chronic conditions; see Table 2 for listing of illness combinations focusing on nine primary condition types: heart conditions, hypertension, diabetes, stroke, mental health, thyroid conditions, lung/pulmonary conditions, osteoporosis, and arthritis.
Sample demographic characteristics (n = 38).
Chronic condition order and combinations of sample (n = 38).
Procedures
The authors’ Institutional Review Board approved all data collection procedures (IRB #: Pro2015000290). To recruit participants, a referral list was obtained from the Director of the NJISA geriatric clinic that included patients known to have at least two chronic illnesses based on medical records, and no record of cognitive impairment/dementia. Recruitment letters describing the study were sent to a list of 201 patients who met inclusion criterion. The letter indicated that they would be contacted by a research team member in the coming week. Of the 201 referred, 42 were reached by the study staff and screened for eligibility, of whom 38 were screened as eligible and completed the interviews. Reasons for non-eligibility included not self-identifying as having two or more chronic conditions and a lack of interest in participation after completing the screener.
Participants provided written informed consent to participate. A semi-structured interview was pilot tested and administered to participants by eight trained medical school and undergraduate student research assistants at participants’ homes (n = 23) or in the NJISA research office (n = 15), depending upon participant preference, during the summer of 2015 (June to September). Multiple interviewers were trained to ensure that one researcher’s perspectives and biases did not systematically influence participant narratives and results. Interviews were audio-recorded, transcribed verbatim, and lasted an average of 64 min (range: 29–135 min). Participants were compensated with a small gift.
Measures
Demographic characteristics. Participants reported their age, education from 0 (no formal education) to 9 (doctoral degree), race as 0 (non-Hispanic white) or 1 (African-American), gender as 0 (male) or 1 (female), marital status as 0 (married ), 1 (widowed), or 2 (single), and annual income from 1 (less than $15,000) to 6 (greater than $150,000) at the time of their interviews.
Timeline. Participants completed a Chronic Illness Timeline with interviewers at the start of their interview. Respondents were told that: Chronic illnesses are conditions like diabetes, arthritis, and depression that last for a long time, require treatment, and are diagnosed by a doctor or other health care professional. They were instructed not to include illnesses such as colds, flu, or surgeries unless these were directly related to a chronic illness. Each illness was entered on the Chronic Illness Timeline along with age at diagnosis. Example timelines are depicted in Figure 1.
(a) Chronic illness timeline of a 72-year-old female, (b) chronic illness timeline of an 81-year-old male.
Semi-structured interview. This study used a semi-structured interview to understand participants’ experiences developing and managing their MCCs. Questions were asked in a series for each illness on the individual’s timeline, starting with the first illness (condition 1) and doing the same for each subsequent illness (up to condition 8). Questions included: Tell me about your experiences with [Condition]. What did you do when you were told you had [Condition]? Did you change any of your daily habits, such as your diet, exercise, drinking, smoking, or medication use? Tell me about how getting this diagnosis of [CONDITION 1] affected the way you dealt with your [OTHER/PREVIOUS CONDITIONS]. The interviewer used follow-up questions and prompted to encourage respondents to elaborate upon answers.
Data preparation and analysis plan
Following Content Analysis strategies, interviews were transcribed, double-checked, and read by the authors (six individuals) using an open-coding process to produce a base coding-tree that reflected the manifest content of the interviews. 33 The team met weekly to discuss and develop a list of agreed upon core codes following an iterative inductive process of theme development.34,35 This approach resulted in the development of broad coding schemes informed by the research questions that were then expanded as the most salient issues and common responses were identified. Categories were further refined as more specific sub-topics became evident; where lower levels of specificity could not be agreed upon by the team, the broader code was maintained to ensure future replicability of the findings. Codes were developed that captured each unique idea. 33 For example, transcripts were coded for each type of self-management behavior described. Each participant response that followed an interviewer prompt was treated as a unit of coding. The transcripts were then re-read and coded by the team utilizing the developed coding tree in QSR NVIVO 10 (QSR International Pty. Ltd). The research team met weekly, coding 20% (n = 6) of the transcripts together. Three team members coded each of the additional 32 transcripts to avoid the potential for a single researcher’s perspective and assumptions to influence data coding and interpretation. Discrepancies between coders were discussed and resolved by consensus.
The team collectively reviewed final codes and transcripts to identify and interpret key themes in the data. In review of the transcripts, it was evident that each person experienced a unique pattern of onset and progression of illnesses. While this is a critical finding in and of itself (discussed below), coded transcripts were used to expand participant timelines to include changes in self-management behaviors over time as a way to visualize the data. Timelines along with transcript texts were examined for patterns in the application of codes. Data saturation in identified themes was attained, whereby no new codes were presenting upon review of new transcripts.
Results
Older adults’ understanding of the timing and progression of MCCs
No two individuals experienced the same pattern of illness development (see Table 2). Timing varied from age 4 to 90 with onset of illnesses occurring simultaneously to 60 years apart. Figure 1 displays timelines for two individuals who each reported four chronic health conditions, illustrating an example of two different patterns of illness onset, progression, and initiation of self-management behaviors. As depicted in Figure 1(a), this 72-year-old female experienced an onset of her conditions within a four-year period. She first described being diagnosed with emphysema at the age of 68, after which she started a medication and reduced her smoking. Two years later she was diagnosed with hypertension, which resulted in an additional medication and change in diet. This was followed by osteoporosis and arthritis each a year after the other, with new medications for both and an increase in exercise. Meanwhile, Figure 1(b) depicts an 81-year-old male whose conditions accumulated over the course of 30 years. His first illness onset was at the age of 45 with diabetes; this resulted in a new medication, monitoring his diet, and increasing exercise. His second condition, hypertension, was diagnosed 10 years later at the age of 55, for which a new medication was started, and arthritis followed 10 years later with another new medication. At the age of 75, he started to experience neuropathy, which resulted in an additional medication, decrease in exercise, and decrease in ability to drive.
In addition to diversity in the timing of onset and progression of illnesses across individuals, two common themes were present in the older adults’ narratives concerning the timing and progression of MCCs: (1) illness narratives were constructed within the context of a person’s life events and/or health events; and (2) illnesses were not predominantly seen as connected to one another.
Illness narratives were constructed within the context of a person’s life and/or health events. Across the sample, illnesses and implementation of self-management behaviors were described within the context of psychosocial events in each person’s life. In other words, diseases were more often incorporated into one’s life and not the reverse. How a person recalled the timing of their illnesses and/or impact was often related through the telling of a tale regarding a distinct psychosocial or health event. Events were also often described as key determinants in how participants responded to the onset or management of a particular chronic illness. Further, psychosocial life experiences more generally were described as impacting individuals’ interpretation of the severity of illnesses. Thus, diseases were not merely perceived as bodily events that called for behavioral action but were also meaningful experiences that impacted a person’s sense of self and expectations. See Table 3 for text examples of each trend.
Examples of illness narratives constructed within the context of a person’s life.
Illnesses were not seen as connected to one another. A second key finding was that the majority of participants (n = 32 out of 38) did not see their illnesses as connected to one another. As one 77-year-old female indicated: “I don’t see where one deals with the other,” or as another 86-year-old female indicated: “It’s like apples and oranges.” The participants “compartmentalized every one” (72-year-old male) and followed “certain rules” (81-year-old male) for each separate condition. Illnesses were described as “one really [having] nothing to do with the other” (82-year-old female). The lack of described connection across illnesses led to the portrayal of unique narratives for each illness in a person’s life, meaning that personal behavioral responses to the onset of each condition were also seen as distinct for each condition.
In a minority of cases, participants described a perceived link between two of their illnesses; these links reflected views that the two conditions were connected due to common etiology or one illness exacerbating the symptoms of the next (i.e. anxiety or depression got worse when diagnosed with diabetes, lung disease, or Parkinson’s), or that the illnesses were each just “one more in the pile” (86-year-old female). When a connection was perceived, the interpretation of connection was unprompted by doctor interactions and was a self-determined connection. As one participant indicated: “We didn’t discuss the relationship between the other illnesses….I understand that…atrial fibrillation can be a result of a thyroid, of hyperthyroidism…we didn’t discuss, you know, anything about it” (91-year-old female). Or as another described: ‘No one connected the dots. But the congestive heart failure and the breathing is all one thing’ (77-year-old female).
The impact of timing and progression on self-management strategies
In regard to our second research question, regardless of condition or order of conditions, all participants described how they had responded behaviorally to each illness over time. Patterns of types of behavioral response were not evident by illness type or order. Participants described five types of self-management behaviors following onset of a new condition: exercise, diet, addictive substance use, medication use, and other adaptive behaviors (see Table 4 for examples). As demonstrated in Table 4, participants described how and when they implemented each type of change. Some participants described making specific modifications to their behaviors each time an illness was diagnosed, describing an increase or decrease in behavior: “I started riding the bicycle a little bit” (64-year-old male, after heart condition diagnosis) and “I quit smoking” (83-year-old female, after hypertension diagnosis). Others described maintenance of behavior: “I didn’t restrict myself to certain diet. I just keep eating the same thing as I was before” (79-year-old male, after cancer diagnosis).
Self-management behaviors.
For most participants, again, there was a lack of perceived connection between illnesses which resulted in a lack of perceived connection between self-management behaviors across conditions. A few participants did indicate that changes made for one illness also helped another (i.e. changes for diabetes helped arthritis for an 83-year-old female; starting Tai Chi for osteoporosis and arthritis benefited emphysema and hypertension for a 72-year-old female), yet this was not dominant across the sample..
Discussion
The goal of this study was to take a lifespan approach to understanding how older adults narrate their experiences of the progression of MCCs over time and resultant changes in their self-management behaviors. Findings demonstrate that the process by which MCCs unfold in the lives of older adults is complex and unique to each individual. Participants interpreted and described their illnesses within the context of their life experiences and predominantly viewed each chronic illness as a distinct entity, for which his responses were not connected across illnesses. Participants described making changes in self-management behaviors at various points along their illness trajectories upon receiving each illness diagnosis. These results expand the literature and carry implications for research and practice guidelines for care of MCCs.
Prior research has focused on understanding the impact of MCCs on individuals and health outcomes by predominately using total count variables or looking at a small number of illnesses. 10 By taking a lifespan approach to understanding illness onset and progression, we found great variability in how illnesses present, progress, and are perceived. No two individuals shared the same illness timeline and having four conditions did not simply mean four conditions—they were characterized by unique challenges, interpretations, and responses. In considering how older adults described their trajectory of illness development, it was evident that illness narratives were constructed within the context of a person’s life events and/or health events. One’s health was not perceived in isolation from psychosocial experiences; time was significant not just in reference to order of disease but how the experiences of illnesses fit into each person’s life. This is consistent with prior theoretical work that postulates that illness is experienced within the context of psychological, social, and environmental experiences. 17 Second, illnesses within the MCC context were not predominantly seen as connected to one another by older adults. This description by participants is likely a product of the fact that even when individuals have MCCs, treatment guidelines are typically linked to a specific illness. 36
Regarding how older adults describe their engagement in self-management behaviors across a life course of illness development, results indicate that participants describe distinct behavioral responses to the diagnosis of each illness. Given that a key clinical goal is the engagement of older adults in positive self-management behaviors to reduce the burdens of living with chronic illness, 7 these results are critical, indicating that older adults are linking behavioral responses to their illnesses. However, not addressed here are the motivations by which older adults act and change their self-management behaviors following diagnosis of each illness. Prior work indicates that the way a person perceives his illness impacting his life impacts his motivation for self-management behaviors. 37
Strengths and limitations
Overall, this work is strengthened by its use of ground-up, qualitative research methodologies that allowed the voice of the participant to guide all analytic inferences. However, the study is not without limitation. First, generalizability of results is limited to older adults in New Jersey who reported two or more chronic health conditions, received care from a single geriatric clinic, were English speaking and able to participate in an hour-long, open-ended interview. Individuals receiving care at NJISA are unique in that they access care from a geriatrician, a physician trained in working with older adults and delivery of individualized care; this may have influenced the experiences of respondents in this study. However, it is worth noting that despite this specialized care, connectivity across illnesses was still not perceived. Further, the majority of participants were white and women. As a result, findings may be restricted to the experiences of these groups of individuals. Prior work indicates that cultural and language differences contribute to low health care literacy which may impact behavioral engagement in care. 38 As a result, more could be learned by studying the processes examined here in a larger, more diverse sample. Second, the reported lifespan timelines represent individuals’ representational narratives of their experiences and ultimately examine the individuals’ perceptions of their past and how they create meaning around their MCCs, as reported at one point in time. Results here may be limited by retrospective recall biases. Examining such processes with a prospective quantitative empirical design that considers current behaviors at the time of illness diagnosis and follows them over time may also shed further light on systematic associations of illness timing, progression, and onset in older adults and the subsequent impact on self-management. Third, this analysis does not consider other factors, such as self-perception of aging, social support, doctor’s support, doctor advice, or access to resources (i.e. financial, healthcare), which have been identified as influential to the experience of managing chronic conditions. In particular, there may be significant others in an older person’s life that serve as key informants and/or supports in managing care 39 or impact outcomes such as pain or mobility. Understanding the way doctors support patients with MCCs and the specific advice provided, and/or specific advice recalled by patients, may further inform uptake of health behaviors. Or, the timing of care-seeking by participants and/or access to financial resources to access care may have determined the timing of diagnosis and therefore impacted response to illnesses. Further work should consider these factors in understanding the timing and progression of MCCs in older adults’ lives. Similarly, interview prompts did not include discussion of illnesses such as the cold, flu, surgeries, or the experience of injuries and recovery from injuries. These additional health-related events may impact the development and management of MCCs and should be examined in future work.
Implications
These findings uniquely highlight the need for person-centered care for individuals with MCCs—honoring a person’s current symptoms but also history and life experiences. Given that no two individuals shared the same trajectory of illness development, there is a need for practitioners to avoid disease-specific guidelines in providing treatment and rather consider the unique holistic experience of the individual. Specifically, if practitioners are attempting to ignite behavioral change, our findings suggest that it is the person’s history and life experiences that will shape how a person behaviorally responds to a diagnosis. We set up a framework by which practitioners can investigate the circumstances impacting individual patients’ behaviors, asking from a lifespan perspective about timing and progression of illnesses. Obtaining specificity regarding each individual’s timeline would allow practitioners to individualize treatment plans and monitoring strategies. Furthermore, from a research perspective, a rudimentary count of illnesses fails to account for context and individuality. For example, Pruchno and colleagues 40 found that it is the specific combinations of illnesses that are linked to depressive symptoms and not merely the number of conditions. Our findings suggest that the timing of onset and progression of illnesses within the context of life experiences adds critical information to our empirical and clinical conceptualization of living with MCCs.
The findings regarding the lack of perceived connection across illnesses carry particular implications for engagement in self-management. If individuals fail to see their conditions as connected, individuals may engage in one behavior for one illness that is counter to recommendations for another illness. Complex self-care regimens may be further seen as unattainable when managing each condition individually. 41 Furthermore, this lack of perceived connection has implications for building evidence-based person-centered care treatment plans as well; there is a need to consider perceptions of connection in making recommendations for behavioral change. The narratives provided by our participants demonstrate a need for practitioners to discuss recommended treatment responses to illnesses in tandem or holistically across illnesses to maximize well-being and care outcomes, and for health systems to redesign care procedures to truly address multimorbidity42—such as communication across specialties and development of systematic patient education programs that address multimorbidity. Furthermore, although clinical practice predominantly focuses on the response to one or two conditions in a single visit due to time and resource constraints of practitioners, this work suggests a need to expand visit time and/or visit structure to accommodate discussion of more conditions simultaneously to provide truly person-centered care.
In addition, given that individuals are able to identify behavioral responses aligned with the diagnosis of each illness over time, probing questions and use of motivational interviewing techniques by clinicians 43 may help determine how and why behavioral changes are initiated, stopped, or maintained over time. Knowledge of prior response to illness experiences and motivation for change may be leveraged to guide individuals to act similarly or differently in response to current or future illnesses.
Conclusions
Findings from this study form an initial empirical understanding of how older adults narrate their development of MCCs over time and how older people describe self-management behaviors along the way. Taking an individualized, holistic approach to care for individuals with MCCs that takes into account individuals’ past development of illnesses, life circumstances, and behaviors is critical to advancing the quality of care provided to older individuals living with MCCs. Additional research that quantitatively, prospectively explores the impact of the development of MCCs within the context of life experiences on self-management behaviors in older adulthood is critical to further substantiate these findings and clarify clinical recommendations for care.
Footnotes
Acknowledgements
The authors would like to thank the older participants who agreed to give their time and the research assistants who supported data collection efforts including Ommul Ali, Christopher Carey, John Polunin, Aashiki Shah, and Nicole Sica.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Summer Medical Research Fellowship and Osteopathic Heritage Foundation Fellowship granted to medical school students at Rowan University School of Osteopathic Medicine.