Adaptation and validation of the Treatment Burden Questionnaire (TBQ) in English using an internet platform

Treatment burden is defined as the ‘work’ of being a patient and its effect on the quality of life (QOL) of patients. It represents the challenges associated with everything patients have to do to take care of themselves (e.g. medication intake, drug management, self-monitoring, visits to the physician, laboratory tests, lifestyle changes, administrative tasks to access and coordinate care) [1–3]. This work can represent a tremendous investment of time, attention, cognitive energy, and effort. For example, a patient with a chronic condition may spend 2 hours or more per day on health-related activities [4], and this increases for patients with multiple chronic conditions who receive treatment or recommendations for each of their diseases [5, 6]. For these patients, treatment burden could be considered a crucial outcome for disease management [7]. Difficulties with their treatment are often not shared in depth by patients during medical consultations [8], and physicians are often not aware of the challenges their patients face in coping with everything asked of them [9]. Thus, a tool that is simple to understand and easy to administer is needed to identify overburdened patients in daily practice and for research, in order to design new therapeutic strategies that are both efficient and acceptable for patients.

The Treatment Burden Questionnaire (TBQ) is the only existing instrument that measures treatment burden without restricting its scope to a single condition or treatment context [9, 10]. The TBQ is composed of 13 items rated on a Likert scale ranging from 0 (not a problem), to 10 (big problem). The TBQ was derived from a literature review and qualitative semi-structured interviews with patients in France. It assesses the burden associated with taking medicine, self-monitoring, laboratory tests, doctor visits, need for organization, administrative tasks, following advice on diet and physical activity, and social impact of the treatment. Item scores can be summed into a global score, ranging from 0 to 130. The instrument was first developed with a sample of patients with one or more chronic conditions, who were recruited from hospitals and general-practice clinics in France.

To measure treatment burden in English-speaking countries, the TBQ needs to be adapted: items must remain comprehensible without changing their original meaning, and it should be recognized that new items could arise in different contexts whereas some existing items might be irrelevant. Patients must be involved in the process to ensure that the resulting tool is clear and relevant [11].

In this study, we aimed to assess the validity and reliability of an English version of the TBQ intended to be used in both routine practice and in research for patients with at least one chronic condition living in the USA, Canada, UK, Australia, or New Zealand.

Adaptation and validation of the TBQ in English followed a multi-step approach recommended in the literature [12–14]: 1) review the conceptual evidence about treatment burden in English-speaking countries; 2) translate the TBQ into English; 3) pretest the instrument with patients to assess the relevance of items, clarity and wording and; 4) assess validity and reliability by a test–retest of the adapted instrument.

We found a high completion rate (>99%) for all items. All items were relevant for all patients, with the proportion of ‘does not apply’ ranging from 0.6% to 7.2%, with the exception of the burden associated with self-monitoring (30.3%) (see Additional file 2). The TBQ global score was the sum of the answers to each of the 15 items, and ranged from 0 to 150 (‘Does not apply’ was considered the lowest possible score). The global score had a high correlation with the score for each item of the TBQ (r_s = 0.47 to 0.71) (see Additional file 3).

Factor structure, assessed by scree plots, favored a unidimensional instrument. The first factor explained 87% of the variance, and had an eigenvalue of 5.91. For subsequent factors, eigenvalues wereless than 0.80 (see Additional file 4).

Reliability was assessed by a test–retest method for 282 patients (46%). Participants in the retest showed similar characteristics as patients in the test phase (Table 1). For the TBQ global score, the ICC for agreement was 0.77 (95% CI 0.70 to 0.82) (see Additional file 6). Bland and Altman plots for the global TBQ score showed a mean difference of −0.58; 95% limits of agreement −43.5 and 42.3 (Figure 1).

Subgroup analysis showed a high TBQ global score for patients who needed equipment such as wheelchairs or canes (mean ± SD score 69.0 ± 33.4), those who received physical therapy (62.7 ± 33.9), and those who had gastrointestinal (65.4 ± 32.5) or skin (64.9 ± 30.8) diseases (see Additional file 7). TBQ scores did not differ for the most common healthcare location (public hospital, private hospital, general practice, or specialist clinic) (P = 0.97).Overall, treatment burden scores were homogeneous and comparable between countries. The lowest mean (SD) TBQ score was for the item related to self-monitoring (from 0.6 ± 1.1 in Canada to 1.7 ± 2.8 in Australia/New Zealand) and the highest was for the item related to the effect of healthcare on relationships with others (from 4.0 ± 3.5 in Canada to 6.7 ± 3.8 for Australia/New Zealand) (Figure 2).

We adapted the TBQ for patients with one or more chronic conditions living in the USA, Canada, the UK, Australia, or New Zealand. This is the first valid and reliable instrument assessing treatment burden across multiple conditions and treatments.

During the adaptation process, we added two new items to the English TBQ that did not exist in the original study, which took place in France [9]. First, we added an item about the financial treatment burden. In France, the national health program guarantees healthcare free of charge for patients with chronic conditions. However, in countries where the English TBQ was administered, this is not always the case [10, 17]. For example, 30% to 47% of patients in the USA may have problems paying medical bills [32]. Second, during the pretest, patients expressed difficulties that they had in their relationships with healthcare providers. This item was not included in the original French questionnaire because the authors felt that it measured a concept different from treatment burden. However, it was mentioned in subsequent qualitative studies [2, 15, 16] and, in the present study, participants insisted on including this item to take into account how discussion with physicians could affect how much they knew about their conditions and treatment, and how non-optimal relationships with healthcare providers could hinder adherence to treatment, lifestyle changes, or follow-up. During the first study, patients may have felt hindered in mentioning this issue to the investigators during face-to-face interviews and therefore the point was underestimated. Use of a web-based questionnaire has shown a decrease in socially desirable responses, and an increase in willingness to disclose sensitive information [33].

Our results highlight the advantages of using an internet platform: we were able to include 200 patients with chronic conditions from different countries in the pretest in less than 1 month. Therefore, we gathered rich qualitative data on questionnaire items and possible missing concepts, indicating the high content validity of our tool.

The item properties of the English TBQ were comparable with those from the original study in France with 1) a significant floor effect (>15%) for all items, which reflects that patients could have no burden in the aspects of their care that they have already integrated into their lives, and 2) about 30% of ‘Does not apply’ for the item related to self-monitoring. As for the original questionnaire in French, we chose to keep the item about self-monitoring because it was considered a mandatory item for other patients, particularly those with diabetes.

The validation study included 610 patients from different countries with different conditions and treatments. As hypothesized, we highlighted statistically significant associations between the overall treatment burden and adherence to medication, patients’ confidence in their knowledge about their conditions and treatment, and quality of life. To our knowledge, this is the first study to report such associations independent of disease or treatment contexts. Development of interventions to improve patient knowledge about therapeutic options could lead to shared decision making, reduced treatment burden,better adherence to medication, and better quality of life for patients.

We compared TBQ scores between countries and found that most item scores were homogeneous for patients in the USA, UK, Canada, and Australia/New Zealand, with two exceptions. First, we found increased financial treatment burden for patients residing in the USA and Australia. This result was expected because qualitative studies in these countries described the financial burden as one of the issues most widely discussed by patients [10, 17]. Second, in the UK, we found increased treatment burden related to patients’ relationships with their healthcare providers. This finding could be related to pay-for-performance initiatives (for example, Quality and Outcomes Framework) that may encourage physicians to focus on biomedical objectives and spend less time listening to patients’ non-medical problems. More research is needed to understand the precise aspects of treatment burden, and why patients might experience different aspects in each of these countries.

This study has several limitations. First, to measure patients' confidence in their knowledge about their treatments and conditions, we used questions that were not validated. However, participants reviewed these questions during the pretest in the same way they reviewed the questions of the TBQ. Patients expressed no difficulties in answering these questions, and we found neither any missing answers, nor a floor or ceiling effect. Second, we recruited patients through the PLM website only, and from this patient group, only 20% of the invited patients responded to the web survey. This response rate was similar to that from other studies involving online surveys. This recruitment method might have selected patients willing to share their experiences with others and/or with severe or rare conditions. We did not include participants who were not able to access a computer. Therefore, our sample is not representative of the general population of patients with one or more chronic conditions: overall patients in our study were younger, more educated, and more often female (similar to other studies involving the PLM website [34–36]). Prevalence of chronic conditions in our study differed from those of chronic conditions of patients seeking healthcare in national surveys. For example, we found a high proportion of patients with neurological conditions, because PLM was initially developed for these patients. Although these biases were considered relatively small, psychometric properties for the TBQ (especially reliability estimates) might differ in other populations. Despite these limitations, use of an online platform has a number of advantages. First, we were able to involve a larger number of patients than in usual cross-cultural adaptation studies (800 participants including pretest) [37] in a very short time (2 months from pretest to results). Second, the recruited patients were geographically dispersed and thus, we might have avoided selecting only patients living near research centers. Finally, patients involved were engaged in research, and therefore provided us with rich feedback at every stage of our research, which may have contributed to the development of questionnaire items closer to patients’ own words and problems.

The present study advances the evidence of the validity of the TBQ. Further studies should focus on testing other psychometric properties such as responsiveness in longitudinal and prospective studies. The TBQ can be implemented in clinical trials assessing interventions intended to mitigate treatment burden and its negative effect on quality of life to develop a Minimally Disruptive Medicine [1].

We adapted and validated the TBQ for English-speaking countries (USA, UK, Canada, and Australia) using an online patient-based platform for direct interaction with patients. This study resulted in the first instrument assessing treatment burdenfor any condition and treatment context in these countries.