How Well Are We Performing the Initial Assessment of HIV-Positive Patients?

Results From a Multicentre Cohort in Spain

I Suárez-García; B Alejos; E Delgado; M Rivero; JA Pineda; I Jarrin


HIV Medicine. 2020;21(2):128-134. 

In This Article


In this multicentre Spanish cohort, we have identified some aspects of the initial evaluation of HIV-infected patients that are not performed adequately. The Spanish AIDS Study Group (GeSIDA) published a list of quality indicators for the care of persons infected with HIV, along with the standards showing the minimum proportion of patients for which each quality indicator was achieved.[8] For the initial evaluation of the variables analysed in our study, the standards of care are measurement of CD4 count, viral load and renal function in 100% of patients, analytical examinations (including hepatitis A, B and C serology, syphilis serology, and total cholesterol measurement) and information on smoking status in 95%, and cardiovascular risk assessment (including blood pressure measurement) and latent tuberculosis screening in 90%. In our cohort, only the examinations of CD4 count, viral load, and, for later years, total cholesterol and creatinine achieved or came close to the standard of care. We found a clear trend for increasing frequency of the examinations of hepatitis A and syphilis serologies, blood pressure and smoking status over the study period, but all these examinations were still well below the standard of care. The frequency of latent tuberculosis screening was consistently low during the whole study period.

The multivariable analysis showed that, after adjusting for other risk factors, patients who acquired HIV by use of injected drugs or by heterosexual transmission had a significantly higher risk of not having most of the examinations performed in the initial assessment. Patients with a low education level also had a higher risk of incomplete assessment for a substantial proportion of the examinations. These findings suggest that these groups have an incomplete initial assessment and raise concerns about the perceived comorbidity risk of these patients as determined by their health care workers and about their access to and retention in the health care system.

Patients of non-European origin had a better initial evaluation for hepatitis A and latent tuberculosis screening, probably because physicians perceived a higher risk of infection among patients from countries with a higher prevalence of these infections. Patients > 50 years old had a higher probability of having their blood pressure determined (probably reflecting concerns about cardiovascular risk in elderly persons). Patients with CD4 counts < 200 cells/μL had a lower probability of being screened for latent tuberculosis, which is particularly concerning as this is the group with a higher probability of tuberculosis reactivation.[9] Physicians might not perform tuberculosis screening in these patients because they might presume that the tests are highly likely to be negative as a consequence of immunosuppression; in fact, although the Spanish and American guidelines recommend tuberculosis screening in all patients,[2,3] the European guidelines only recommend it in patients with CD4 counts > 400 cells/μL.[1]

This is, to our knowledge, the first study that has specifically evaluated the initial assessment of HIV-infected patients in a large multicentre cohort and that has investigated factors associated with having an incomplete initial evaluation. A previous study evaluated several quality indicators in nine Spanish hospitals which included some of the items assessed in our study, with very similar results: initial CD4 count and viral load were performed in 98% of the patients, and hepatitis C serology, cardiovascular risk assessment and latent tuberculosis screening were performed in 86%, 62% and 43% of the patients, respectively.[4]

Our study has several strengths: a large multicentre cohort with strict quality control, and which is highly representative of the Spanish population.[10] Our study shows how research cohorts may be used to evaluate adherence to clinical guidelines, showing the advantages of readily available data and large numbers of patients assessed, which might not be achievable with other clinical cohorts or review of clinical records. There is, however, one limitation: as we were analysing routine data, we cannot exclude the possibility that some of the examinations could have been performed but not recorded in the database. However, we believe that the proportion of such examinations was probably low, because: (1) some of the examinations were recorded in a high percentage of patients, (2) we found consistent trends over time, and (3) the results are consistent with the previously published work cited above.[4]

In conclusion, although the initial assessment was adequate for CD4 count and viral load, it was still suboptimal for the evaluation of cardiovascular risk and smoking status, syphilis and viral hepatitis screening, and diagnosis of latent tuberculosis. Although we found consistent trends for improved assessment in more recent years, some examinations were still far from the quality standards. Efforts should be made to better evaluate all patients at their initial visits, and in particular vulnerable groups such as injecting drug users and persons with a lower education level, who had a higher risk of having an incomplete initial assessment. Monitoring the completeness of the initial evaluation, and giving feedback from this monitoring to clinical staff, could potentially increase adherence to the clinical guidelines and improve quality in the health care of HIV-infected patients.