Patient perceptions about illness self-management in ANCA-associated small vessel vasculitis

Abstract

Objectives

To characterize patient perceptions, related to eight self-management behaviours relevant for adults with ANCA-associated small vessel vasculitis (ANCA-SVV), and to determine if these perceptions were associated with performance of each behaviour.

Methods

Adults with ANCA-SVV (n=202) completed a self-administered questionnaire that assessed eight self-management behaviours (adherence to recommendations for medication, health service use, diet, exercise, infection avoidance and symptom monitoring; prompt reporting of symptoms and side effects; and adjusting activities in response to symptoms), perceptions about these behaviours, socio-demographics, clinical factors and social desirability bias. Descriptive statistics were generated to characterize patients’ perceptions about difficulty of, importance of, and specific barriers to performing each behaviour. Regression analyses explored whether these variables were associated with performing each behaviour, controlling for potential confounders.

Results

With few exceptions, higher perceived importance and lower perceived difficulty of each behaviour were associated with more frequent performance of the behaviour. For each behaviour, several specific barriers were frequently endorsed by patients and a number of these were associated with lower levels of self-management.

Conclusion

This study reveals that patient perceptions about the illness and its treatment influence ANCA-SVV self-management. Perceived barriers to medication, health services, diet and exercise adherence were similar to those in other illnesses. This study also provides insight into barriers experienced by patients in performing behaviours (infection avoidance, symptom monitoring, reporting symptoms and side-effects and adjusting activities) not often previously studied. How the identification of these barriers can help inform future interventions for ANCA-SVV patients is to be discussed.

Introduction

ANCA-associated small vessel vasculitis (ANCA-SVV) is a group of rare autoimmune conditions, characterized by inflammation and necrosis of blood vessels primarily in the respiratory tract and kidneys [1]. ANCA-SVV is potentially fatal, and has an often progressive but unpredictable course [2]. Treatment with immunosuppressive medication induces remission in 80–100% of patients [3], but risk of relapse and morbidity from the disease and its treatment remains high [4].

In order to minimize these risks, individuals with ANCA-SVV are required to engage in illness self-management, defined as the ‘day-to-day tasks an individual must undertake to control or reduce the impact of disease on physical health status’ [5]. This involves adhering to physicians’ explicit behavioural recommendations (e.g. taking medication, following a diet), making informed decisions about care, and applying skills to maintain adequate psychosocial functioning [5]. In ANCA-SVV, behavioural recommendations may be numerous, come with considerable risk of side-effects [6], and require substantial role adjustment and lifestyle changes [2]. However, self-management of ANCA-SVV has very rarely been researched. This is concerning, given that poor adherence to behavioural recommendations is common across chronic illnesses [7] and in the case of ANCA-SVV, could lead to complications such as renal failure [8], reduced quality of life or death.

A multitude of studies have identified many factors that influence self-management in more common illnesses [7]. This research and behavioural theory suggest that patients’ perceptions about the illness and its treatment, including perceived barriers to performing recommended behaviours, may determine self-management behaviour [7, 9–11]. Other perceptions, including beliefs about illness severity or potential outcomes, benefits of the behaviour, susceptibility to complications and confidence in performing the behaviour, are also widely believed to affect illness self-management [10]. However, this research also shows that levels of adherence and barriers to performing recommended behaviours vary across specific behaviours [7, 11–13]. Furthermore, a recent review underscored the importance of developing self-management intervention strategies that are tailored to the demands of specific illnesses [11]. In order to develop effective strategies for improving self-management in ANCA-SVV, we must first understand which behaviours are most challenging for patients, and identify factors that influence their performance.

Thus, our first aim was to characterize patients’ perceptions about ANCA-SVV self-management, including eight behaviours that are relevant for most patients. We examined three specific questions: (i) How difficult is it for patients to perform each behaviour? (ii) How important is it to patients to perform each behaviour? (iii) What specific barriers to performing each behaviour do patients perceive? Our second aim was to determine if these perceptions were associated with the frequency with which patients actually performed each behaviour.

Patients and methods

We conducted a cross-sectional, observational study, in which participants completed a self-administered, mailed questionnaire. Participants received $10 for participation. This study was approved by the Biomedical Institutional Review Board of the University of North Carolina.

Sample

Participants had to be ≥18 yrs, be able to read and write in English, and report a diagnosis of ANCA-SVV (e.g. WG, microscopic polyangiitis, Churg–Strauss syndrome, ANCA-glomerulonephritis). Recruitment involved sending letters to patients enrolled in the Glomerular Disease Collaborative Network’s vasculitis disease registry, and making announcements in patient newsletters and websites and at support group meetings. We identified and mailed consent forms to 275 eligible and interested patients; 230 (84%) provided consent and 205 (75%) completed the questionnaire. Three additional patients who reported an ineligible disease type (Goodpasture’s syndrome, n=2 and temporal arteritis, n=1) were excluded from this analysis, for a final sample of 202 patients.

Measures

Self-management behaviour

ANCA-SVV self-management was measured using the Vasculitis Self-Management Scale or VSMS [14], which assesses the frequency, on a 5-point Likert scale, with which patients perform up to eight self-management behaviours: (i) medication adherence, (ii) adherence to recommended health services (attending appointments and obtaining medical tests and immunizations), (iii) infection avoidance adherence, (iv) dietary adherence, (v) exercise adherence, (vi) symptom monitoring adherence, (vii) adjusting activities in response to fatigue or symptoms and (viii) prompt reporting of new or increased illness symptoms or side-effects. For the first six behaviours, respondents report how frequently (e.g. 1=none of the time to 5=all of the time) they performed the behaviours as recommended by a health professional; accordingly, they only complete the items if a health professional has recommended the behaviour. For each behaviour, they also complete items providing additional detail on providers’ behavioural recommendations (e.g. number of medications prescribed). For the remaining two behaviours, which are generally recommended for all patients, respondents again report how frequently they performed the behaviour, but all respondents complete the items unless they have not experienced any symptoms or fatigue during the recall period. In line with convention [15], a recall period of the past 4 weeks is used for daily behaviours (e.g. medication-taking), and the past year for less-frequent behaviours (e.g. attending appointments). Several items are negatively worded and reverse-scored, and higher scores correspond to higher levels of self-management.

Specific barriers to self-management

For each behaviour, participants also reported, using a checklist, the factors they believed most frequently or consistently got in the way of them performing the behaviour during the recall period. Each checklist was developed from in-depth interviews conducted with 18 patients and was pilot-tested with eight additional patients as part of the development of the VSMS [14] to ensure that it was appropriate and comprehensive. The content and number (range 8–20) of barriers in each checklist varied for each behaviour. Respondents were also permitted to write in additional barriers not included in the checklists. Dichotomous variables (1=barrier endorsed, 0=barrier not endorsed) were then created for each barrier in the checklist or listed by participants. Because we originally expected items in the VSMS about attending appointments and obtaining tests or immunizations to factor into separate behavioural domains [14], perceived barriers to these behaviours were assessed separately in this study. The same was true for reporting symptoms and reporting side-effects.

Perceived difficulty and importance of behaviours

Respondents rated how difficult (1=‘not at all’ to 6=‘extremely’) and how important (1=‘not at all’ to 6=‘extremely’) it is for them to perform each behaviour exactly as recommended. These items were also pilot-tested as described above and demonstrated comprehensibility and variation in responses. As with specific barriers, perceived difficulty and importance were assessed separately for attending appointments and obtaining tests or immunizations, and for reporting symptoms and reporting side-effects.

Socio-demographic and clinical variables

Patients reported their age, gender, race/ethnicity, years of education and marital status. They also reported their history of dialysis and kidney transplant and illness duration, and rated how active their vasculitis currently was (1=‘not at all/remission’ to 10= ‘extremely’).

Social desirability bias

Social desirability bias was assessed with the short form of the Marlowe–Crowne Social Desirability Scale [16]. This 20-item scale (Cronbach’s α=0.77) assesses whether survey responses are influenced by respondents’ desires to appear in socially desired ways.

Analysis

Analyses were conducted using SAS 9.13 (SAS Institute, Cary, NC, USA). Descriptive statistics for all variables were examined, including perceived difficulty and importance ratings for each behaviour, and the frequency with which specific barriers to each behaviour were endorsed. Due to the large number of specific barriers in each checklist, we report only those endorsed by at least 10% of the respondents.

To assess whether perceived difficulty and importance predicted patients’ self-management behaviour, a series of regression analyses were conducted, with separate models for each behaviour. Each model included the perceived difficulty and importance variables, along with eight control variables (gender, white race or not, age, years of education, married or not, illness duration, current disease activity and social desirability bias). For the outcome of adherence to recommended health services, perceived difficulty and importance of attending appointments with health professionals as well as obtaining medical tests and immunizations were analysed as independent variables in separate regression models. Likewise, for the outcome of prompt reporting of illness symptoms and side-effects, perceived difficulty and importance of reporting illness symptoms as well as reporting treatment side-effects were analysed as independent variables in separate regression models. Thus, a total of 10 regression models were examined in this set of analyses.

Because these analyses involved the evaluation of 20 regression coefficients (i.e. 10 models × 2 primary independent variables per model=20 coefficients), we used the Benjamini–Hochberg procedure for controlling the false discovery rate [17, 18], which controls the expected proportion of false rejections for the set of comparisons to <α/2 [18, 19]. In this procedure, the P-values associated with the 20 β-coefficients were listed in ascending order and compared with linearly interpolated critical values ranging from α/2 to α/2/m (where m=20, or the total number of comparisons, and α=0.05). Observed P-values that were lower than their corresponding critical values were deemed statistically significant.

We conducted another set of regressions to assess whether perceptions of specific barriers were associated with self-management behaviour. Each model consisted of one of the eight self-management behaviours as the dependent variable, a specific barrier as the independent variable and the eight control variables listed earlier in the article. We again used the Benjamini–Hochberg procedure to evaluate the statistical significance of each test within this set of regressions (m=55, α=0.05).

Results

Descriptive characteristics

Sample characteristics are shown in Table 1. About half of the sample was female, and almost all respondents were white. The mean age was 55 yrs, and the mean education level was 15 yrs (i.e., 3 years of college). The majority reported a diagnosis of WG, and the mean illness duration was 76 months, or just over 6 yrs (median=52.0 months or 4.3 yrs).

Table 1.

Descriptive statistics for ANCA-SVV patients (n=202)

	n	Percentage or mean (S.D.)
Socio-demographics
Gender
Male	93	46.0
Female	109	54.0
Race/ethnicity
White	188	93.1
American Indian/Alaskan Native	2	1.0
Asian	5	2.5
Black or African American	2	1.0
Hispanic/Latino	4	2.0
Missing	1	0.5
Age (yrs)	202	54.9 (14.6)
Years of education	202	14.6 (2.4)
Marital status
Married	152	75.3
Non-married	50	24.8
Clinical characteristics
Self-reported condition
WG	145	71.8
Microscopic polyangiitis	16	7.9
Churg–Strauss syndrome	10	5.0
ANCA-glomerulonephritis	31	15.4
Time since diagnosis (months)	201	75.8 (70.9)
Patient-perceived disease activity, range 1–10	201	2.9 (2.1)
Ever on dialysis
Yes	30	14.9
No	169	83.7
Missing	3	1.5
History of kidney transplant
Yes	14	6.9
No	184	91.1
Missing	4	2.0

Information provided by patients with regard to specific behavioural recommendations received from providers revealed that all self-management behaviours assessed in the VSMS were relevant for a majority of patients. Almost all patients had been instructed to take self-administered medication during the past 4 weeks (95%), attend appointments with health professionals in the past year (99%), and obtain immunizations or medical tests in the past year (99%). Patients were taking a mean of 6.7 (s.d.=4.1, range 0–18) medications. The mean number of health professionals that patients had been recommended to see in the past year was 4.4 (s.d.=2.0, range 0–11), and nephrologists (67%), primary care physicians (67%), opthalmologists (54%) and rheumatologists (51.5%) were the most commonly seen. In addition, a majority had been instructed to take steps to avoid getting an infection (69%), follow dietary recommendations (78%), follow exercise recommendations (61%) and monitor illness symptoms (65%). The most common infection avoidance behaviours recommended to patients included staying away from people who are ill (64%) and washing hands frequently (61%). The most common dietary recommendations included following a low-salt diet (40%), limiting alcohol intake (37%) and trying to lose weight (36%). Among exercise recommendations, general cardiovascular exercise (e.g. walk most days of the week) was most commonly recommended (58%). Among symptom monitoring tasks, at-home blood pressure (43%) and weight (38%) monitoring were most common. Over 90% of patients reported feeling tired, ill or run-down during the past 4 weeks at least ‘a little of the time’, and thus completed items on adjusting activities in response to fatigue or symptoms. Roughly 93% and 85% of respondents completed items on reporting symptoms and side-effects, respectively.

Descriptive statistics for frequency of performance of each self-management behaviour are shown in Table 2. Paired t-tests revealed that levels of adherence to recommended health services and medication were significantly higher than all other types of self-management (all Ps<0.0001), and that levels of exercise adherence and prompt reporting of symptoms and side-effects were significantly lower than all other types of self-management (all Ps<0.05).

Table 2.

Descriptive statistics for VSMS subscales^a, including mean frequency of performance of each self-management behaviour

	n	Mean (S.D.)	Cronbach’s α
Medication adherence	188	4.5 (0.54)**	0.77
Adherence to recommended health services	199	4.6 (0.63)**	0.79
Infection avoidance adherence	139	3.8 (0.79)	0.85
Diet adherence	151	3.6 (0.76)	0.78
Exercise adherence	123	3.1 (1.01)†	0.94
Symptom monitoring adherence	131	3.8 (0.97)	0.91
Adjusting activities in response to fatigue and symptoms	182	3.4 (0.65)	0.67
Reporting symptoms and side-effects	174	3.2 (1.08)†	0.90

^aRange of all eight VSMS subscales was 1–5.

^**Statistically greater than all other types of self-management (P<0.0001).

^†Statistically lower than all other types of self-management (P<0.05).

Perceived difficulty and importance

Summary statistics for perceived difficulty and importance ratings are shown in Table 3. Mean difficulty ratings were lowest for medication-taking and obtaining tests and immunizations, and highest for exercise, diet and adjusting activity levels in response to fatigue or symptoms. Mean importance ratings were highest for medication-taking, attending appointments and obtaining tests and immunizations, and lowest for exercise, diet, adjusting activities, reporting symptoms and reporting side-effects. Within each self-management behaviour, Pearson correlations of perceived difficulty and importance were all negative and ranged from quite small (r=−0.06 for adjusting activities) to moderate (r=−0.42 for exercise).

Table 3.

Summary statistics for perceptions about specific ANCA-SVV self-management behaviours

	Perceived difficultya (Mean s.d.)	Perceived importancea (Mean s.d.)
Medicationb	1.5 (1.0)	5.9 (0.4)
Appointments with health professionals	2.0 (1.4)	5.7 (0.7)
Tests and immunizations	1.8 (1.3)	5.7 (0.6)
Adjusting activity level	3.0 (1.6)	4.8 (1.3)
Reporting symptoms	2.4 (1.5)	4.8 (1.4)
Reporting side-effects	2.2 (1.5)	4.7 (1.5)
Infection avoidance	2.7 (1.7)	5.4 (0.8)
Diet	3.0 (1.5)	4.8 (1.2)
Exercise	3.2 (1.5)	4.7 (1.2)
Symptom monitoring	2.0 (1.3)	5.1 (1.1)

^aRange of perceived difficulty and importance scales was 1–6.

^bPerceived importance and perceived difficulty were assessed globally, for all medication regardless of mode of administration (i.e. self or health professional).

Table 4 shows results of regression models examining the relationship of perceived difficulty and importance to each behaviour. Perceived difficulty ratings of specific behaviours were independently and negatively associated with patients’ reports of their performance of that behaviour, across all behaviours and perceived importance was positively related to each behaviour.

Table 4.

Relationship of perceived difficulty and importance to the frequency with which patients performed vasculitis self-management behaviours

		Perceived difficulty			Perceived importance
	n	β	s.e.	P	β	s.e.	P
Medication (self-administered)	184	−0.20	0.03	<0.0001*	0.38	0.07	<0.0001*
Appointments with health professionalsa	185	−0.11	0.03	0.0016*	0.15	0.06	0.0240*
Tests and immunizationsa	194	−0.19	0.03	<0.0001*	0.36	0.05	<0.0001*
Infection avoidance	136	−0.26	0.03	<0.0001*	0.28	0.06	<0.0001*
Diet	148	−0.24	0.03	<0.0001*	0.24	0.04	<0.0001*
Exercise	121	−0.36	0.06	<0.0001*	0.28	0.07	0.0001*
Symptom monitoring	130	−0.39	0.05	<0.0001*	0.29	0.07	<0.0001*
Adjusting activity level	179	−0.17	0.03	<0.0001*	0.15	0.03	<0.0001*
Reporting symptomsb	170	−0.34	0.05	<0.0001*	0.25	0.05	<0.0001*
Reporting side-effectsb	170	−0.34	0.04	<0.0001*	0.23	0.05	<0.0001*

^aDependent variable was adherence to recommended health services.

^bDependent variable was prompt reporting of illness symptoms and side-effects.

^*P-value is significant, according to Benjamini–Hochberg procedure.

Specific barriers to self-management

A total of 55 barriers to performing the various self-management behaviours were reported by at least 10% of patients. The number of barriers for the specific behaviours ranged from 1 (obtaining tests and immunizations) to 11 (exercise adherence). Based on the Benjamini–Hochberg procedure, 34 of the 55 barriers were significantly associated with the relevant behaviour. In all cases, the associations were negative; thus, presence of the barriers was associated with less frequent performance of the relevant behaviour. A comprehensive listing of all 55 barriers and the frequency with which they were endorsed, as well as their statistical relationships (unstandardized β-coefficients, standard errors and P-values) to the respective self-management behaviour, is available (see Supplementary Table 5, available as supplementary data at Rheumatology Online); those that were significantly related to behaviour are discussed subsequently.

Five perceived barriers were associated with lower medication adherence. The first two of these barriers represented disruptions to patients’ daily routines (β=−0.40, s.e.=0.08) and forgetting (β=−0.58, s.e.=0.07), while the remaining three factors pertained to the complexity of the medication regimen (large number, β=−0.30, s.e.=0.11; difficult instructions, β=−0.42, s.e.=0.11; complicated dosing schedule, β=−0.43, s.e.=0.12). Perceived barriers to adherence to recommended health services represented logistical issues in scheduling and getting to appointments (e.g. conflicting work responsibilities, trouble scheduling appointments at convenient times, long travel times), but none were related to adherence. Two perceived barriers were associated with lower infection avoidance adherence: patients’ desire to participate in social/leisure activities (β=−0.32, s.e.=0.14) and honour work responsibilities (β=−0.70, s.e.=0.15). Four perceived barriers were significantly associated with worse dietary adherence: frequent exposure to prohibited foods (β=−0.42, s.e.=0.13), competing food preferences (β=−0.61, s.e.=0.13), emotional eating (β=−0.64, s.e.=0.15) and low motivation to adhere (β=−0.59, s.e.=0.16). For exercise adherence, five perceived barriers were associated with worse adherence: lack of energy (β=−0.51, s.e.=0.20), lack of motivation (β=−0.99, s.e.=0.17), competing work (β=−0.74, s.e.=0.23) and family (β=−0.68, s.e.=0.24) responsibilities and disliking exercise (β=−1.16, s.e.=0.24). Two perceived barriers were associated with worse symptom monitoring adherence: forgetting (β=−0.99, s.e.=0.15) and a belief that the behaviour was not necessary for controlling symptoms (β=−1.13, s.e.=0.25). With regard to appropriate adjusting of activities, patients who less frequently cut back on their activities when they felt tired, ill or run-down were more likely to report the following barriers: competing family (β=−0.26, s.e.=0.10) or work (β=−0.34, s.e.=0.11) responsibilities, not feeling comfortable with asking others to help with responsibilities (β=−0.42, s.e.=0.11) and feeling that others would not understand (β=−0.41, s.e.=0.11). Finally, a number of barriers predicted less prompt reporting of illness symptoms and side-effects: believing the symptom (β=−0.93, s.e.=0.17) or side-effect (β=−0.89, s.e.=0.16) would go away on its own, uncertainty as to whether the symptom (β=−1.04, s.e.=0.16) or side-effect (β=−0.68, s.e.=0.16) was vasculitis-related, desire to wait to report the symptom (β=−1.17, s.e.=0.15) or side-effect (β=−0.90, s.e.=0.17) at the next appointment, not wanting to bother the health professional about the symptom (β=−0.93, s.e.=0.17) or side-effect (β=−0.85, s.e.=0.18), difficulty in reaching a health professional about the symptom (β=−0.73, s.e.=0.20) or side-effect (β=−0.76, s.e.=0.21), not wanting to be prescribed more medication as a result of reporting a new symptom (β=−0.85, s.e.=0.26), and the physician not explicitly recommending prompt reporting of symptoms (β=−0.75, s.e.=0.27).

Discussion

To our knowledge, this is the first study to examine patients’ perceptions about ANCA-SVV treatment regimens and their relationships to self-management behaviour. Our results suggest that patients believe that among the self-management challenges they face, they experience the least difficulty in adhering to recommended health services and medication. Patients reported the highest levels of adherence, the highest levels of perceived importance and lowest perceived difficulty for these behaviours. In contrast, analysis of these same indicators revealed that patients appeared to have the most difficulty with exercise adherence, adjusting activities in response to fatigue or symptoms and prompt reporting of symptoms and side-effects. Patients also perceived dietary recommendations as more difficult and less important to follow, although mean levels of actual dietary adherence were more moderate. These findings are not surprising, given that prior research with other illness populations has shown that patients are typically less adherent to self-management behaviours involving lifestyle modifications, such as diet and exercise [11, 12, 20].

As expected, the more important patients perceived a specific self-management behaviour to be, the more frequently they reported performing it as recommended. These results are not surprising in light of prior research among individuals with other serious chronic illnesses (e.g. HIV, breast cancer) that have shown patients’ beliefs about the necessity and health benefits of performing recommended behaviours to be related to their actual performance of the behaviour [21, 22]. Thus, perceived importance of specific vasculitis self-management behaviours may be one promising avenue for future interventions. For example, patients who do not believe it is very important to limit their salt intake could be presented with information on possible consequences (e.g. kidney failure and the need for dialysis or transplant) of not doing so. Clinicians may also want to be explicit about the reasons why it is important that patients carry out their behavioural recommendations when they are discussed in office visits.

In general, patients who perceived a behaviour to be more difficult tended to perform that behaviour less frequently. Beyond this expected result, this research identified specific factors that patients commonly believe make performing vasculitis self-management behaviours more difficult. These barriers varied substantially across behaviours, as did significant relationships between reporting of these barriers and actual behaviour. This provides further support for the argument [11] that each self-management behaviour is distinct, and efforts to improve patients’ level of self-management should be behaviour-specific.

Within each self-management behaviour we assessed, patients’ reporting of a number of specific barriers was found to significantly predict their level of performance of the behaviour. These perceived barriers may be especially fruitful avenues to explore in future research and intervention with ANCA-SVV patients. While space limitations prohibit an exhaustive discussion of each of these factors, a few findings warrant highlighting here.

Among the behaviours that have been extensively studied with other patient populations, including adherence with regard to medication, health services, diet and exercise, many of the most frequently reported barriers in this study have been previously documented as influences on adherence. For example, one recent study found that adults with type 2 diabetes who believed that changes in their daily routine made it more difficult to take medication had lower levels of adherence to hypoglycaemic medication [23], just as vasculitis patients reported in this study. Forgetting and regimen complexity, which were associated with medication non-adherence in this study, are also widely reported as common reasons for non-adherence in other illnesses [24]. As in other rheumatic disease populations [25], a high proportion of patients in this study also indicated that the difficulty with which they could schedule appointments at preferred times played a key role in their ability to adhere to recommendations for using health services, although reporting of this as a barrier was not related to self-reported adherence. Also, geographic distance from health care providers was also frequently mentioned as a barrier to attending appointments with health professionals, just as in other patient populations [26]. Likewise, the specific barriers mentioned by patients and associated with worse dietary adherence, namely frequent exposure to prohibited foods, food preferences, emotional eating and low motivation to adhere, have been found to be powerful predictors of dietary behaviour in other populations [27, 28]. Finally, lack of motivation, time and facilities for exercise have all been associated with reduced physical activity levels in prior research [29] and are reflected in this study as well. These results suggest that researchers and clinicians may want to look to interventions that have been used successfully with other populations for guidance in developing intervention strategies for ANCA-SVV patients to improve adherence in the domains of medication-taking [24], appointment-keeping [30], immunizations [31, 32], diet [28] and exercise [29]. For example, providing patients with pillboxes to organize their medications may decrease forgetting and barriers due to complexity of the regimen, and has been shown to be related to increased adherence in other chronic conditions [24, 33, 34].

Much less research has been conducted on infection avoidance adherence, symptom monitoring adherence, adjusting activities in response to fatigue or symptoms and prompt reporting of symptoms and side-effects. In this study, the barriers reported for avoiding infections revealed that patients’ desire to carry on with their normal activities, whether they are work-related or social/ leisure-related, conflicted with their adherence. With regard to adjusting activity level in response to fatigue or symptoms, respondents once again commonly perceived that their work and family responsibilities made it difficult to obtain the rest their bodies needed. Many respondents also reported reluctance in reducing their activities or asking for help because of worries about others’ reactions to them doing so. Given that previous research has found that a substantial proportion of individuals with ANCA-SVV experience disruptions at work [4, 35], as well as reductions in household income and negative influences on interpersonal relationships due to their illness [4], it is not surprising that some patients are hesitant to reduce their activity levels or delegate their responsibilities to others despite signals from their bodies that they should, or to reduce risk of infection. There is also some evidence that quitting work due to ANCA-SVV may actually negatively impact health-related quality of life [35]. Our results, considered with this prior research, suggest that with regard to these behaviours, the costs of strictly adhering to selfcare guidelines may outweigh the benefits. Much work is needed to develop and test ways to improve adherence in these realms that will be acceptable to patients and not negatively impact their quality of life.

With regard to symptom monitoring adherence, patients who reported forgetting and a belief that it was not necessary for controlling one’s condition as barriers had lower adherence. Interventions that focus on routinizing symptom monitoring and stressing the importance of close monitoring, such as Herlyn and colleagues’ patient education programme for German primary systemic vasculitis patients [36], may be warranted. One approach that may be easily implemented in clinical practice is for physicians to encourage patients to create a blood pressure monitoring ‘cue’ in their daily routine; e.g. place the blood pressure monitor on a bedside table so that patients are more likely to remember to take their blood pressure when they get into bed at night. This type of behavioural cue has been shown to improve routinizing and adherence with regard to medication [24], and may also facilitate home monitoring of illness symptoms.

Finally, prompt reporting of new or increased illness symptoms and side-effects appears to be a particularly difficult issue for many patients, as demonstrated by the frequency with which a number of barriers to this action were endorsed by patients. Most notably, patients reported that their uncertainty as to whether symptoms are related to vasculitis or are meaningful enough to report often served as barriers to prompt reporting. Many patients were wary of going through the trouble of trying to speak with a health professional or ‘bothering’ the health professional about something that might not be important, and were inclined to wait and see if the symptom would resolve on its own. Future studies should further explore the role of the patient–physician relationship and specific provider behaviours, as barriers or facilitators to prompt reporting of symptoms and side-effects by ANCA-SVV patients. Interventions that address patients’ concerns as well as factors on the provider side may be warranted. For example, our results suggest that providers’ active encouragement of patients to contact them immediately with any symptom or side-effect may facilitate more prompt reporting. In addition, providing patients with multiple mechanisms to directly contact the physician (e.g. e-mail in addition to phone or pager), may also be effective, although research is needed directly in testing the efficacy of these efforts.

Several limitations of this study should be noted. First, our design was cross-sectional, which limits our ability to draw causal inferences about observed relationships. Second, we used a convenience sample that may not represent the larger population of ANCA-SVV patients. Our sample may over-represent patients with kidney involvement, as well as those with WG vs microscopic polyangiitis or Churg–Strauss syndrome. Furthermore, patients’ average disease duration was over 6 yrs and recently diagnosed individuals are under-represented. However, the socio-demographic and clinical characteristics of our sample were similar to those of other US studies with systemic vasculitis patients [4, 37]. Third, our measurement of perceived barriers was less than ideal in that we used single items created specifically for this study, with unknown validity. However, these items were pretested with ANCA-SVV patients and found to be comprehensive and appropriate. Finally, this study was limited to assessing patients’ perceptions of barriers, and should by no means be considered an exhaustive study of factors that impede ANCA-SVV self-management. Respondents may not explicitly recognize the factors that influence their behaviour.

Despite these limitations, this study provides insight into ANCA-SVV patients’ perceived barriers to performing a number of self-management behaviours. Identification of these barriers should be useful in informing more formal study of important influences on illness self-management in the context of ANCA-SVV. Furthermore, our detailed identification of barriers to the performance of eight specific health behaviours relevant to managing ANCA-SVV should facilitate the development of new interventions supporting self-management specifically in ANCA-SVV patients. Depending on the behaviours of interest, this may involve developing entirely new strategies, adapting existing self-management programmes shown to be effective in other chronic illnesses, or enhancing the only known vasculitis-specific patient education intervention, conducted in Germany, [36] to specifically address the behaviour-specific barriers identified in this study. This study serves as a starting point for this work.

Rheumatology key messages.

• Perceptions about self-management predict health behaviour in ANCA-SVV patients.

• Specific barriers that patients report for each behaviour, including several unique to ANCA-SVV, can help inform future self-management interventions.