Procedure

The study was conducted in three phases:

Development of the adherence questionnaire

Validation of the developed adherence questionnaire

Assessment of adherence in relation to personality traits and achieved goals using the developed questionnaire

An “item-pool” consisting of 46 items was generated based on literature on adherence in combination with our experience and knowledge from our previous research (Table 1) [12, 14, 19, 20]. A 46-item questionnaire was constructed that began by explaining the aim of the questionnaire in an accepting tone. In order to reduce the risk of social desirability bias, it has been recommended that questions about adherence be posed in a way that presents missed doses as something normally occurring [21]. The respondent was asked to take a position on each statement by choosing among five response alternatives (“Doesn’t correspond at all”, “Doesn’t correspond very well”,”Neither corresponds nor doesn’t correspond”, “Corresponds fairly well”, and “Corresponds exactly”) on a 5-point Likert-type scale [22]. The order of the items was constructed so that the questionnaire did not begin with controversial or emotive items. There was also a mixture of positive and negative statements, which is suggested as a method of minimizing response bias [23].

The item pool. Items marked 1 and 2 were used in psychometric tests in phase 1 and 2

The questionnaire was sent by mail to 300 adult individuals with asthma using asthma medication, selected at random from the population-based West Sweden Asthma Study (WSAS). The following is a brief explanation of the WSAS. In 2008, 30,000 randomly selected individuals were invited to take part in the WSAS, by postal questionnaires, and 18,087 completed and returned the questionnaires [24]. Out of the 18,087 individuals from the general population who completed the questionnaires, 2006 individuals, including 964 adult asthmatics, participated in the following clinical phase of the WSAS . The respondents in the current study have been selected from the asthma cohort (n = 964), who reported that they had used asthma medication during the past 12 months. In 2013, they were invited to participate through one initial mailing of the questionnaire and one reminder. A completed and returned questionnaire was regarded as consent to participate in the current study.

Following psychometric analysis of the questionnaire in phase 1, a new questionnaire was constructed, similar in design to that used in phase 1. However, one item was slightly changed from “I don’t follow the doctor’s prescription exactly but instead I can feel what I need to take” to “I don’t follow the doctor’s prescription of asthma medicine exactly but instead I can feel how much I need to take.” The change was made in order to clarify that the item was supposed to assed the self-adjusting of the doses, which was not clear in the origin item “what I need to take” and to emphasize that it was asthma medication that was in focus.

In 2013, 200 adult individuals with asthma using asthma medication were invited to take part in phase 2 of the current study i.e. the validating of the developed adherence questionnaire. They were randomly selected from the asthma cohort in the clinical phase of the WSAS. They had not been included in phase 1 described above. They received the adapted questionnaire and, in addition, the Medication Adherence Report Scale (MARS) [25]. One initial mailing of the questionnaire and one reminder were sent. A completed and returned questionnaire was regarded as consent to participate in the study.

The sample in phase 3 consisted of the same respondents as in phase 2 described above. The respondents completed the following questionnaires on personality traits and goals with the asthma medication:

In order to assess personality traits, the Neuroticism, Extraversion and Openness to Experience Five-Factor Inventory (NEO-FFI) consisting of 60 items, scaled 1–5, was used [15].

In order to assess goals with the asthma medication treatment the following two questions were used:

“What is your goal with your asthma medication?” This was an open question and the respondents stated their personal goals with the asthma medication.

“Do you think that your goal with your asthma medication has been reached?” This question was used to assess whether the stated goal with the asthma medication was reached or not. Yes and no were used as response alternatives.

The study samples were described using descriptive statistics, i.e. frequencies, percentages, means and standard deviations [26]. Factor analysis, i.e. principal component analysis with Varimax and Kaiser normalization as the rotation method, was used to explore associations between the items, to remove redundant items and to identify underlying constructs among the items [27]. Cronbach’s alpha was used to test the reliability of the scales in the developed adherence questionnaire [26]. In order to demonstrate validity, Pearson’s correlation coefficient was used to determine associations between responses to the new questionnaire and the MARS (in phase 2) [23]. Associations between adherence and personality traits were investigated using Pearson’s correlation coefficient and a multiple regression model. T-tests were used to study differences between subgroups. All statistical analyses were performed using SPSS version 20. Reports on goals with the asthma medication were compiled into categories according to content.