A Survey of Patients’ Perspectives on Outpatient HIV Care in the Netherlands

In The Netherlands, HIV outpatient care is provided exclusively in acknowledged HIV treatment centres (n = 26 at the time of this study). The care is generally provided and coordinated by specialised clinicians (predominantly infectious disease specialists with experience in HIV medicine) and specialised nurses (“nurse consultants”). All treatment centres collaborate with the collection of data for the ATHENA observational cohort and thus all HIV-infected patients in care (with exception of those who opt out, 2%) are registered and monitored in the cohort database [9].

We conducted a cross-sectional study evaluating patients’ preferences and experiences in a sample of people with HIV in outpatient care in The Netherlands between July 2013 and December 2014.

A nationwide sample of potential participants was selected from the ATHENA database. All HIV-1-infected patients in outpatient care on 25 February 2013, aged 18 or older at time of diagnosis and using cART for at least 6 months, were eligible for participation. We randomly selected samples from each HIV treatment centre in The Netherlands, 1000 patients in total, ensuring a minimum of 20 patients per care facility and accounting for the total number of patients in care in each centre (larger samples in larger treatment centres). We used the pseudonymised study ID number assigned to each HIV patient in care in the cohort to select potential participants.

During their next outpatient visit selected patients were approached for participation by the nurse consultants, who provided them with an information letter and a password for accessing an online survey in Dutch or English. We also offered a paper version of the survey. The Medical Ethics Review Committee of the Academic Medical Centre of the University of Amsterdam exempted this study from written informed consent. We informed all patients about the study and considered consent implicit when a survey was completed online or a hard copy was returned to us.

Data regarding age, sex, region of origin, socioeconomic status (SES), route of HIV transmission, time since initiation of cART and time since HIV diagnosis were extracted from the ATHENA cohort database.

Region of origin, grouped into The Netherlands, sub-Saharan Africa (SSA) and other, was based on the country of birth. We used an area code-based classification system previously described by The Netherlands Institute for Social Research to determine patients’ SES [10]. Here, we recoded the five classes as high, middle or low. We grouped route of transmission into men who have sex with men (MSM), heterosexual contact or other/unknown. The last category includes intravenous drug use, as this route of transmission accounts for only a very small proportion of HIV infections in The Netherlands [11].

The Netherlands Institute for Health Services Research (NIVEL) previously developed a series of disease-specific questionnaires that measure “Quality of Care through the Patient’s Eyes” (QUOTE) (in, amongst others, asthma, rheumatic disease and frail elderly) [12]. In QUOTE questionnaires, both patient preferences and experiences are assessed. To measure preferences, patients weigh various care aspects by rating whether statements are “not important”, “quite important”, “important” or “of greatest importance” (e.g. “The specialist should always explain the advantages and disadvantages of any treatment”). Subsequently, experiences are measured by letting respondents indicate whether they “always”, “usually”, “sometimes” or “never” receive care as stated (e.g. “My specialist explains the advantages and disadvantages of any treatment”).

The HIV-specific version of this questionnaire (QUOTE-HIV) was developed in 1999 using focus groups (patients), tested for its psychometric properties [8] and subsequently tested in a pilot study in the USA [13]. Our aim was to keep adjustments of the instrument to a minimum, but we modified the original QUOTE-HIV questionnaire by removing six statements that we considered no longer applicable (e.g. regarding inpatient care and euthanasia) or difficult to interpret. We also added five statements (concerning the attitude of the care provider, flexibility in care and referral to other disciplines) based on literature review and suggestions of the Dutch Association for people living with HIV. Our final version consisted of 25 statements regarding care provision by the HIV specialist, 21 statements regarding care provision by the nurse consultant, 5 statements regarding general aspects of care and 5 scores (0–100) on specific aspects of care.

We tested for differences in characteristics of respondent and non-respondent characteristics, using t tests, Mann-Whitney U tests and χ ² tests as appropriate.

For each statement, we calculated the proportion of respondents who considered it “important” or “of greatest importance”, hereafter jointly referred to as “highly preferred”. We then, among the patients who rated the items as “highly preferred”, calculated the proportions of patients reporting “always”, “usually”, “sometimes” or “never” receiving such care.

Across the entire instrument, the proportion of respondents rating items as “highly preferred” (i.e. “important” or “of greatest importance”) ranged from 37 to 98%. Figure 1 presents a selection of the items that were most often rated as “highly preferred”, i.e. by 90% or more of the respondents. The items are grouped by type of care provider and in order of priority. The bars show whether the respondents reported their experience as “always”, “usually”, “sometimes” or “never” receiving care as stated.

The aspects of care that were most often considered “highly preferred” were having an HIV specialist and nurse consultant with specific expertise in the field of HIV and safeguarding the privacy of HIV status. Other themes that stood out concerned how the care provider treats the patient (taking the patient seriously and a sympathetic/involved attitude) and adequate provision of information regarding treatment (advantages and disadvantages, how to take the medication and possible side effects).

Least often viewed as “highly preferred” were aspects regarding waiting times (37% and 59% for the medical specialist and nurse consultant respectively), being aware of the patient’s situation at home or at work (45% and 62%), discussing the possibility of requesting a second opinion (56% and 62%) and not being disturbed during a consultation with the medical specialist (55%).

A number of themes were considered important specifically for the nurse consultant. These themes were: being accessible by phone (94%), working in close cooperation with other care-givers (93%) and taking sufficient time to talk to the patient (92%).

Overall, most patients who rated items as “highly preferred” also reported “usually” or “always” getting such care (85–100%). This applies to all the items in the questionnaire. The rates of reporting “sometimes” or “never” receiving care as stated were somewhat higher in statements regarding the provision of information. Specifically, experiences regarding providing information on advantages and disadvantages of treatments and side effects were somewhat less positive. The median scores that respondents gave to their medical specialists and nurse consultants were 93 and 94 respectively (on a scale of 0–100).

With regard to priorities in general aspects of care, three themes were covered. The referral to other professionals by HIV care providers was rated as “highly preferred” by 85% of respondents, who gave a median score of 85 for this item. Being able to schedule an outpatient visit at a suitable time was rated “highly preferred” by 73%, and 92% of these patients reported usually or always being able to make an appointment at a suitable time. Extensive opening hours of the department for blood draws was considered a “highly preferred” item by 57% of respondents, and its median score was 82. Respondents gave the waiting times at the department for blood draws a median score of 76.

The strength of this study is that we have data from a national sample of HIV-infected patients in which each of the designated HIV care facilities in the country is represented. In addition, we used an HIV-specific instrument that was developed together with patients and that measures preferences as well as experiences. Furthermore, to reduce potential sampling bias related to online survey research [17], respondents were provided with the option of filling out a hard copy version of the questionnaire.

There are however also important limitations to take into account. Despite our efforts to recruit a nationally representative sample of HIV-infected patients, people from The Netherlands, MSM and patients with a higher SES were overrepresented in the respondents. Moreover, the response rate in this study was relatively low (35%) [12, 18], with rates differing considerably across the centres. There is a possibility that patients’ experiences impacted their willingness to respond. Indeed, previous studies have found an association between more favourable patient perceptions and higher participation rates in health survey studies [18]. The fact that respondents were more likely to have an undetectable viral load supports this notion. However, evidence also suggests that the overall impact of non-response bias may be small and that increasing participation may not change conclusions of patient perception surveys [19, 20]. In addition, our adjustment for SES has its limitations because it is based on area codes and not on individual patient data (e.g. education and occupational status). In the absence of a validated instrument applicable to the current era of HIV care, we used an instrument that was developed more than a decade ago and has not been used extensively. Furthermore, since the development of this instrument much progress has been made in terms of methods for measuring important aspects of the patient care experience. However, in view of the fact that our instrument was developed systematically and jointly with patients and covers HIV-specific themes that are still relevant, we believe our results have yielded useful information.

Our results underline the importance of providing information that meets the needs and expectations of patients. Receiving adequate information has emerged as a particular area of concern in many previous studies [21‐24]. Our results therefore confirm that those involved with the delivery of care should continuously be engaged with which information patients wish to receive, how this information should be provided and whether the provision indeed meets patients’ needs.

In addition, our data show that confidentiality of patients’ HIV status is of great importance in clinical practice. Practices should consider assessing patients’ specific concerns so that measures to minimise confidentiality breaches can be identified. Examples of such issues include the physical environment (e.g. separating the reception area from the waiting area), staff-to-staff contact and the manner of ‘calling the patient into consultation’.

Respondents reported positive experiences across all aspects of care covered by the questionnaire. Given that the HIV care system in The Netherlands has a long history of being assigned to providers and facilities with HIV-related experience, these positive results may reflect an overall good quality of HIV care. However, considering that the questionnaire was developed in the beginning of the cART era, the questionnaire may have failed to capture a number of themes that are important to the current HIV-infected population [4]. Acknowledged areas of particular interest may include psychosocial support [25, 26], patients’ involvement in medical decisions (i.e. ‘shared decision making’) [1, 7] and peer support [26]. Other topics that are worth investigating are patients’ opinions on relocation of services (including the management of non-HIV co-morbidities) from specialised centres to peripheral health facilities or general practitioners and patients’ wishes regarding the frequency and duration of appointments. In the absence of a gold standard, the HIV care field would benefit from a novel and validated instrument to assess HIV-infected patients’ preferences and experiences. Crucial for the development of such an instrument is that patients are involved throughout the process [27]. In addition, the development process should be guided by an established framework. One of the ‘patient-centred care’ frameworks that has been used extensively in research is the Institute of Medicine’s six dimensions of patient-centred care [1, 18]. The dimensions consist of: (1) respect for patients’ values, preferences and expressed needs; (2) information, communication and education; (3) coordination and integration of care; (4) emotional support—relieving fear and anxiety; (5) physical comfort; (6) involvement of family and friends [28].