Hepatitis C and HIV Combined Screening in Primary Care

A Cluster Randomized Trial

Javier Martínez-Sanz; María Jesús Vivancos; Matilde Sánchez-Conde; Cristina Gómez-Ayerbe; Lidia Polo; Cristina Labrador; Patricia González; Alba Mesa; Alfonso Muriel; Clotilde Chamorro; Yolanda de la Fuente; Pilar Pérez Elías; Almudena Uranga; Margarita Herrero; Sara Ares; Rafael Barea; Santiago Moreno; María Jesús Pérez-Elías


J Viral Hepat. 2021;28(2):345-352. 

In This Article


Setting and Design

This is a prospective, cluster randomized, crossover study carried out in four primary care centres of the basic health area of the Ramón y Cajal Hospital in Madrid, Spain, between November 2016 and September 2017 (ClinicalTrials.gov Identifier: NCT03145753). This healthcare area has an estimated population of 555 665. Primary care is distributed between 20 primary care centres, with all of them referring to the same tertiary centre. Four health centres were selected by the research team based on a minimum number of one HIV diagnosis in the previous year along with their desire to participate in the study. The study statistician blindly randomized the centres to either the intervention or the control arm. A cluster randomization strategy was used to facilitate fieldwork and avoid contamination and possible errors due to the large number of professionals and participants that had to be included. The comparison of these centres in terms of population, providers, community and geographic characteristics is provided in Table S1.

Study Procedures

The control arm consisted of a 4-hour educational programme for healthcare providers. The programme included four modules: (a) epidemiology of HCV and HIV infection, (b) diagnostic strategies, (c) treatment of HCV and HIV, and (d) other sexually transmitted infections. It was delivered through a power-point presentation. This educational strategy was recently evaluated by our work group, finding that it can lead to changes in primary care providers' behaviour.

The intervention arm offered the same educational strategy, however, it also included other externally resourced interventions (DRIVE-03 program). They consisted of a 'risk of exposure and indicator condition' (RE&IC) questionnaire, rapid HCV and HIV tests, and five specialized nurses to manage all study procedures. The staff in charge of study procedures received a financial incentive for obtaining the written consents, completing the risk questionnaires, performing the rapid tests and recording all procedures for each participant included.

In the DRIVE-03 screening program, all participants filled out the RE&IC questionnaire. It was previously validated for HIV infection,[19] while its use for HCV is considered exploratory at this time. The validated questionnaire consists of six items that investigate the risk of exposure to HIV (based on identified HIV transmission routes), with 14 items related to HIV indicator diseases (based on a selection of 'HIV Indicator Diseases across Europe Study' [HIDES]).[20,21] In the DRIVE-03 questionnaire, community-acquired pneumonia was added as a new indicator. Any affirmative answer categorized it as positive for HIV risk; however, in the case of HCV, only some of the questions were considered to be risk indicators (Supplementary Material). Participants with at least one positive answer were subsequently tested for HIV or HCV. Moreover, the HCV test was performed on all individuals over 50 years, regardless of the result of the questionnaire. This age group corresponds to the cohort with the highest known prevalence of HCV, and we proposed as a secondary objective to explore whether there was any reported risk factor in addition to age among the participants diagnosed with HCV infection. The anti-HIV 1/2 test WB/S/P (TürkLab Laboratories) was used as a rapid HIV test, and the Anti-HCV WB/S/P test (TürkLab Laboratories) was used for HCV. All individuals with a positive HCV or HIV test were informed and counselled about the results, and the coordination staff in the Infectious Diseases Service of the Ramón y Cajal Hospital was immediately notified. HCV or HIV infection confirmation, which included a complete evaluation within the following 48 hours, was offered to all cases and performed in those who were able to attend.

People aged 18–70 years, attending one of the centres, were offered the option of participating in the study. Exclusion criteria were the presence of a prior HIV diagnosis, and the inability to understand the Spanish language. A sample size of 9072 participants per arm was estimated to detect differences in the rate of new diagnoses, considered as an accomplishment of 1500 rapid tests in each centre. Recruitment was open until inclusion of this number of participants was reached in each intervention centre. Then, the centres crossed to the opposite randomization arm, using the same process. Table S1 shows the time periods, study arm and testing data for each centre. Data on refusal to participate were not collected exhaustively; however, recruitment staff completed a survey about perceived rejections by demographic groups, to estimate their perception of participation based on population characteristics. The reasons for dropping out of the study after it was started were included in the nursing staff worksheets.

The study was approved by the research Ethics Committee at the Ramón y Cajal Hospital and the Madrid Regional Primary Care Research Committee; all participants gave written informed consent. Principal investigators at each centre assisted in the allocation and compliance with all procedures.

Definitions, Data and Statistical Analysis

The main variables considered were: estimated number of people who attended primary care centres and who agreed to participate (obtained through the primary care databases), the number of HCV and HIV tests performed, the screening coverage (defined as the proportion of people screened with the questionnaire of the total number of attendees), and the absolute number and rate of new HCV and HIV diagnoses per 100 000 visits.

The baseline variables collected were: gender, age, country of birth (categorized in Spain, Eastern or Western Europe, Africa, Asia, Latin America and others), the level of the studies (basic, medium or university), the centre of inclusion, result of the RE&IC questionnaire (positive or negative for HIV or HCV) and the result of HIV or HCV rapid tests. In subjects newly diagnosed with HIV, we also collected baseline CD4 + count, adherence to antiretroviral treatment, linkage to and retention of care at 48 weeks, and viral suppression. In subjects diagnosed with HCV, we evaluated genotype, viral load, estimated time from infection, degree of liver fibrosis (measured by elastography), treatment performed (if started) and virologic response at week 12. Data were stored in an Excel-2011 database (Microsoft®), using self-developed, automated questionnaire-reading software. Statistical analysis was performed using Stata 15.1 software (Stata Corp-LP, College Station, TX).

We performed sensitivity analyses without showing significant differences between the study periods. Therefore, the final aggregated data is presented to facilitate the understanding. We used multivariable logistic regression to assess the factors related to a higher probability of presenting a risk questionnaire for both HCV and HIV. All contrasts were bilateral and assumed confidence intervals (CI) of 95%. P values <.05 were considered statistically significant. All analyses were carried out in collaboration with the Biostatistics Department of the Ramon y Cajal Hospital.