Background

The desire to improve the quality of health care has led to the development of assessment tools [1]. Donabedian’s study [2] established the basis of the quality systems applied to health care and defined quality as “an adaptation of the attention to the particular needs of each case”, highlighting that the majority of health care evaluations are limited to a recounting of the actions undertaken, without taking into consideration their effects on health or the extent to which they meet the needs of patients. In recent years, demand has grown for health care that, in addition to being effective and evidence-based, is perceived as satisfactory and beneficial by the patient [3]. This perspective is especially important in patients with chronic diseases, such as infection with the human immunodeficiency virus (HIV), since clinical and physiological measurements provide valuable information for the physician but are often poorly correlated with functional capacity and patient welfare [4]. In the case of patients with HIV infection, the Spanish AIDS Study Group (GeSIDA) published a document in 2010 that includes 66 indicators of care quality. Currently, patients with HIV infection have a longer life expectancy than they did historically and accumulate more comorbidities, which means that health professionals need to address issues beyond virological suppression [5]. User satisfaction is one of the most relevant characteristics that can be used to measure the quality of a service and serves as the basis for the formulation of appropriate health policies [6, 7].

Information reported by patients allows the implementation of actions to improve health care [8]. Some very important elements of health care, such as the perception of the results from the patient’s perspective, are poorly represented in the current health care indicators [9]. In this context, it seems especially relevant to monitor whether there is a relationship between the quality of care achieved through compliance with the established quality standards and the quality of care as perceived by the patient.

The aim of this study was to evaluate the compliance with the quality indicators established by the GeSIDA in a cohort of people living with HIV (PLHIV) and to analyse the influence of those indicators on patient satisfaction. Although some articles have been published on the indicators of the quality of care in the HIV-positive population in different contexts and their relationships with different factors [10, 11], the impact of those indicators on patient satisfaction has not been analysed.

Methods

Study design

This was a cross-sectional observational study of an HIV-positive Spanish population.

Patients

In the analysis of compliance with the quality indicators, all PLHIV over 18 years of age who started or continued follow-up during the study period between September 2011 and November 2017 were included.

For the analysis of patient-reported satisfaction, patients who were not participating in active follow-up at the time of the survey (due to death, transfer to another centre or loss to follow-up), those who were not able to respond (due to illiteracy, language barriers or poor baseline situation), and those who declined to participate were excluded.

The work was carried out in accordance with the Helsinki declaration of 1974 (updated in 2000) and current personal data protection regulations. It was approved by the Research Commission of the study centre and by the Ethics Committee of the hospital. All patients included in the satisfaction analysis signed the patient information sheet and provided informed consent.

Quality indicators

The indicators of the quality of care for HIV-positive patients proposed by GeSIDA were used. There were 66 indicators published in 2010 [12], which were subsequently validated, demonstrating their reproducibility and feasibility [13]. The indicators that all HIV units should monitor to determine their status and identify necessary improvement measures are considered relevant. The indicators that must meet the established standard so that the unit can be accredited are called basic [12]. All the basic and relevant indicators were analysed except for three: indicators 9-Relevant contents of the initial assessment and 33-Periodic consultation report could not be evaluated because the necessary data were not available; neither indicator is considered relevant. Indicator 52-Specific treatment of chronic hepatitis C virus (HCV) was not analysed because during the study period, the treatment of chronic hepatitis C was based on direct-acting antiviral drugs and not on interferon and ribavirin, as stated in the definition of the indicator.

The analysis of compliance with the indicators was carried out by the research team following the definitions established by the GeSIDA experts. The information necessary for the evaluation was obtained from the electronic medical records and was collected in a specifically designed database.

Study variables

To achieve the main goal of the study, patient satisfaction was analysed by means of the survey “External Consultation User Satisfaction Questionnaire” (SUCE), a self-reported previously validated in the Spanish population [14]. It consists of 12 items with a response scale from 1, the worst rating, to 10, the best rating. It allows the independent evaluation of the clinical quality by analysing aspects related to care by health personnel and the administrative quality by analysing organizational and structural aspects. By means of ROC curves, 6.3 was established as a cut-off point to discriminate between satisfied and unsatisfied patients [15].

The questionnaires were offered in English and Spanish at the end of the consultation, and the patients answered them in a separate space, depositing them at the end in an container located in the administrative area of the centre.

Other relevant variables, such as age, sex, educational level, country of origin, date of diagnosis of HIV infection, transmission mechanism, CD4+ cell count and HIV viral load (VL), were analysed.

Statistical analysis

All analyses were performed with the statistical program STATA 14 College Station, TX.

A descriptive analysis of the basic characteristics of all the patients was performed; frequency distributions were used for categorical variables, and means (standard deviations) or medians (interquartile ranges) were used for continuous variables, based on the normality of their distributions. These characteristics were compared between all patients and those participating in the survey with the chi-square test and Student’s t-test to assess the representativeness of the sample.

Forty-seven indicators of the quality of care were evaluated with their 95% parametric confidence intervals (CIs). An indicator was met when the 95% CI of the compliance percentage contained the value defined as the standard.

Linear regression models were used to estimate the differences in means and 95% CIs with regard to the effect of compliance with the indicators on patient satisfaction. For the regression analyses, the indicators calculated at the individual level were included for the patients who participated in the interviews.

To study which quality indicators were independently associated with satisfaction, multivariable regression models were constructed, in which independent variables were included in addition to the indicators age, sex, transmission mechanism, country of origin, CD4+ cell count and VL at the time the questionnaire was completed. Those variables with p <  0.05 were retained in the model.

Results

A total of 334 patients with HIV infection were treated in the outpatient clinic between September 2011 and November 2017. Table 1 shows the descriptive analysis of the patients. A total of 64.1% of the patients were men. Forty-four were foreigners, and sub-Saharan Africa was the most frequent place of origin (24.4%). Fifty percent of the patients had been diagnosed before 2007 (IQR; 1996; 2013), and the most frequent infection risk category was heterosexual (43.2%), followed by parenteral drug addict (PDA) (28.4%) and men who have sex with men (MSM) (25.4%). On the first visit to the centre, the median age was 42 years (IQR: 35; 49), the median CD4+ lymphocyte count was 457 cells/mm3 (CI: 219; 685), and 40.4% had undetectable plasma VLs. The median follow-up time in the centre was 3 years (IQR: 1; 5).

Table 1 Characteristics of patients treated in the HIV program stratified by survey participation
Full size table

To calculate the quality indicators, 334 PLHIV treated in the outpatient clinic during the study period were included. At the time of the survey, 93 patients were not being followed up (lost to follow-up, transferred or died), and 43 patients were excluded due to illiteracy, language barriers, organic or psychiatric pathology that made it impossible for them to complete the questionnaire or refusal to participate. Of the 198 surveys administered, 172 were collected (87% response rate). In the satisfaction analysis, 164 patients were included because 8 surveys were not valid (Fig. 1).

Fig. 1
figure 1

General outline of the study: patients included between 2011 and 2017

Full size image

There were differences between patients included and excluded in the survey analysis. The percentage of participation in the surveys was higher in patients with longer follow-up times (p <  0.001), MSM (p = 0.005) and patients of Spanish nationality (p = 0.038) (Table 1).

Quality indicators and user satisfaction

Compliance with the quality indicators was high: of the 47 indicators evaluated, 35 met the established standard (74.46%). Table 2 shows the results of compliance with each of the indicators. The SUCE scale score was 9.04 (95% CI: 8.90; 9.19) (Fig. 2). Based on the proposed cut-off value of 6.3 [15], 98.16% of the patients were satisfied. The Clinical Quality Factor Score (9.48, 95% CI: 9.37;9 .60) was higher than the Administrative Quality Score (8.56, 95% CI: 8.35; 8.76). Table 3 shows the univariate analysis of quality indicators related to satisfaction (SUCE). Those related to greater satisfaction were 16-Periodicity of visits [difference in means 95% CI: 0.62 (0.13;1.11)], 21-Vaccination against hepatitis A [difference in means 95% CI: 1.14 (0.16;2.12)] and 23-Vaccination against pneumococcal infection [mean difference 95% CI: 0.74 (0.10;1.38).

Table 2 Quality indicators for all patients evaluated
Full size table
Fig. 2
figure 2

Results of the satisfaction questionnaire (SUCE) in the interviewed patients

Full size image
Table 3 Univariate analysis of quality indicators related to satisfaction (SUCE)
Full size table

In contrast, those who met the 26-Evaluation of alcohol intake had a lower level of satisfaction [95% difference in means 95% CI: − 1.20 (− 1.97; − 0.44)]. Foreign patients reported greater satisfaction than Spaniards [95% CI difference: 0.42 (0.12;0.71)] (Table 4).

Table 4 Univariate analysis of other factors related to satisfaction (SUCE)
Full size table

Table 5 shows the multivariable analysis of satisfaction-related factors (SUCE). The final satisfaction model included the indicator 16-Periodicity of visits (regular follow-up) and country of origin.

Table 5 Multivariable analysis of satisfaction-related factors (SUCE)
Full size table

Discussion

The baseline characteristics of the 334 patients treated at the centre were significantly different from those of other Spanish cohorts [17, 18], with a higher percentage of foreigners (40.42%), mainly from sub-Saharan Africa. This may account for the fact that the most common risk category was HTSX and not MSM, as in other published series. Although the male sex predominated, there were more women (36%) than usual in this type of study [18].

Of the 198 surveys administered, 172 were collected, with a response rate of 87%, which is much higher than that published in other works based on self-reported questionnaires [14, 19,20,21].

A low response rate may result in a response bias if the responding patients have significant differences from non-responders [22, 23]. Some authors have proposed that a response rate of at least 80% is acceptable [23, 24], while a response rate of 60% would not be adequate in terms of the representativeness of the sample [23].

Among the patients who participated in the survey, Spaniards accounted for a higher proportion, which is reasonable because the questionnaires were delivered in Spanish and English, and not all foreign patients speak those languages. The authors believe that by providing a separate space in which to complete the surveys and another in which to deliver them, both of which were outside of the consultation room, the participation of the patients was not biased.

Quality indicators

Although the level of compliance with the quality indicators was high, there are areas for improvement. With regard to some indicators, for which the subjective feeling of compliance is high, the result for the indicator was lower than the recommended standard, probably because there is not always a written record in the clinical history of the actions performed, for example, indicators 8-Proof of previous HIV serology, 13-Health education in the initial assessment, 25-Treatment and prevention of smoking and 41-Registration of adherence to treatment.

There is another group of indicators related to the comorbidities of patients with which the level of compliance was suboptimal: 29-Syphilis screening, 49-Evaluation by CHILD or MELD for chronic liver disease, 55-Ultrasound control in cirrhotic patients and 56-Assessment of cardiovascular risk. Due to these results, the performance protocols have been modified, and reminders have been placed in the clinical history to improve compliance.

In the case of indicator 20-Detection of latent tuberculous infection (LTI), which evaluates the performance of LTI screening in the initial assessment, the disagreement between the number of Mantoux tuberculin skin tests requested and performed stands out; the increasingly widespread use of interferon-γ-based techniques in our setting is significantly improving compliance, as it is a technique that is routinely performed with other analyses, which avoids the need for the patient to make two trips to the hospital.

The result of indicator 26-Evaluation of alcoholic intake deserves special mention; the low level of compliance with this indicator is because in daily practice, questions about alcohol intake are asked of patients with a history of excessive consumption and not of all patients during regular follow-up as required in the definition of the indicator.

User satisfaction

The SUCE score was 9.04 points, which is higher than that obtained with the same questionnaire in other Spanish hospitals [7, 25]. It is likely that the higher level of satisfaction is related to the fact that in the hospital conducting this study, each patient is always treated by the same physician. In satisfaction studies conducted in outpatient clinics of different specialties, patients have suggested that to improve satisfaction, they should be attended by the same doctor in successive visits [19, 20].

Based on the proposed cut-off value of 6.3 [15], 98.16% of patients were satisfied at the time of the interview. The low proportion of dissatisfied patients makes comparisons between satisfied and dissatisfied patients difficult. The Clinical Quality Score was higher than the Administrative Quality Score, which is a finding previously described by other authors who have worked with this questionnaire [7, 26].

Quality indicators and user satisfaction

An association was found between the performance of check-ups and analyses at least every 6 months and patient satisfaction; this finding should be interpreted with caution because it is possible that patients who are more satisfied with health care more regularly keep their scheduled appointments and that compliance with this indicator is the result of greater satisfaction and not cause of it.

It was also found that patients who had been vaccinated against hepatitis A virus (HAV) and pneumococcus were more satisfied, but those who had received a recommendation for vaccination against hepatitis B virus (HBV) were not. This disagreement suggests that although statistically significant differences were detected, their clinical relevance is probably minimal.

Compliance with the alcohol intake assessment indicator was associated with less satisfaction because, as it was applied in this population, it identifies patients with extreme alcohol consumption habits. Excessive alcohol consumption is associated with worse patient-reported satisfaction [27, 28].

In this cohort, compliance with indicator 11-Determination of plasma VL at the first visit was 100, and 88.14% of the patients had undetectable VLs when they responded to the survey, although having an undetectable HIV VL did not influence the level of satisfaction.

Some authors have analysed the complex relationship between quality management strategies and the satisfaction perceived by the patient and did not find any significant associations [6]. This fact can be explained by the low impact that the strategies implemented to increase the quality of care have on patient perceptions [6]. The direct personal relationship of the patient with the physician or the nursing staff is a powerful predictor of the patient’s perception [29, 30]. To comprehensively assist patients with HIV infection, care indicators should focus not only on scientific evidence and clinical practice guidelines but should also promote the best clinical practice in other areas, such as organizational aspects, doctor-patient relationships, patient safety and medical errors [9].

Strengths and limitations

The quality indicators used in this study have been validated in the Spanish population, and have been shown to be reliable, with interobserver concordance levels greater than 95% [13].

Some studies evaluated compliance with quality standards and related their results to social, demographic, cultural and clinical factors, but no studies have contrasted these results with the level of satisfaction of the PLVIH as a user of the healthcare system. This perspective provides a more comprehensive evaluation of the quality of healthcare offered to patients.

The satisfaction questionnaire used was validated in the Spanish population and was found to have good psychometric properties. Its use enhances the reliability and validity of our results, unlike other results obtained with non-validated questionnaires [19, 20]. The high response rate also adds value, reducing the possibility of biases and increasing the representativeness of the sample. As it is a generic questionnaire for application in outpatient clinics, it was not possible to analyse aspects specific to HIV infection, such as the type of antiretroviral therapy (ART) used, ART dosage, etc.

Offering the questionnaires in English and Spanish may have led to participation bias because some patients were not able to read in either of those languages. Delivering them immediately after the physician’s evaluation allowed the patient to recall their most recent impression and improved the quality of the information collected. The timing of the survey avoided the recall bias that would be expected if the administration of the survey had been delayed and also prevented the patient’s perception from being influenced by experiences in other areas of the hospital. However, the possibility of a Hawthorne bias, which is the bias that occurs if healthcare staff members modify their habitual attitudes because they are aware that they will be evaluated, must not be overlooked; this limitation is inherent in this type of work. The small number of dissatisfied patients in our study limited the possibility of comparing satisfied and dissatisfied patients.

Conclusions

This study demonstrates the need to analyse our clinical practice with regard to the care of HIV-positive patients to identify areas of improvement and increase the level of patient satisfaction with the care received.

HIV VL is used to measure the effectiveness of ART and is important for the patient’s health but is not related to their perceptions of their satisfaction with the health care they have received. The achievement of the health objectives proposed by scientific societies does not imply the fulfilment of patient expectations. High-quality health care requires that health care professionals satisfy both official requirements and patient needs.