Analysis of the Consequences of the COVID-19 Pandemic on the Continuum of the Care in a Cohort of HIV-Positive Patients Followed in a Single Center of Northern Italy.

Background: During the COVID-19 pandemic, hospitals faced increasing pressure, where HIV patients risked to either acquire SARS-Cov-2 and to interrupt the HIV continuum of care. Methods: retrospective, observational study. We compared the numbers of medical visits performed, antiretroviral drugs dispensed and the number of new HIV diagnosis and of hospitalizations in a cohort of HIV patients followed by the Spedali Civili of Brescia, between the bimester of the COVID-19 pandemic peak and the bimester of October-November 2019. Results: We observed a reduction in the number of new HIV diagnosis from 6.4 in 2019 to 2.5 per month in 2020, a raise of missed visits from 5% to 8%, a drop in ART dispensation especially for not-protease inhibitors and an increase of hospitalized HIV patients due to COVID-19. Conclusions: Our experience highlights the importance of a resilient healthcare system and the need to implement new strategies in order to guarantee the continuum of HIV care even in the context of emergency.

made urgent and mandatory to institute an emergency response aimed to maintain HIV care delivery, adapting our out-patients clinic to the new and dramatic changes imposed by the pandemic.
Here we describe the continuum of HIV care in our patients, the changes made in order to guarantee medical assistance during these two months and the consequences of the pandemic in our patients in these critical months.

Study setting
The present retrospective, observational study was conducted at the University Department of Infectious and Tropical Diseases of the Azienda Socio Sanitaria Territoriale (ASST) Spedali Civili of Brescia, Lombardy, Northern Italy. The Spedali Civili General Hospital of Brescia is one of the largest hospitals in Italy, and it is also the reference hospital for the School of Medicine of the University of Brescia. It represents the only tertiary referral center for Infectious Diseases and HIV care of the whole Brescia province (about 1.2 million inhabitants).

Study population
All HIV-positive patients aged 18 years and more followed at our out-patients clinic were included.
From the administrative les, we collected data on i) new HIV diagnosis ii) number of medical visits in our HIV out-patients clinic (scheduled and then performed trough in-person assessment or telemedicine, missed or postponed iii) dispensation of antiretroviral (ART) medications and iv) hospitalizations of HIV-positive patients. From the electronic and paper clinical records, we retrieved epidemiological and clinical characteristics including sex, origin country, CD4 + T-cells absolute count, HIVRNA load, antiretroviral therapy (ART), diagnosis of admittance to hospital whether hospitalized.
Normally, chronic HIV-positive patients with stable and e cient ART (HIVRNA < 20 cp/ml for at least six months) and CD4 + T-cells > 500 cell/µl are evaluated through blood exams and subsequent medical visit every six months at our clinic. Naïve patients, as well as patients with more complicated medical issues which require a tighter follow-up, are seen more often (6). Antiretroviral drugs are dispensed behind medical prescription every three months by our pharmacy, inside the hospital. During the COVID-19 epidemic, we modi ed the normal HIV management as explained in Fig. 1. Criteria for telemedicine are reported in Table 1. The study comprises two periods: one before and one during the pandemic. In order to describe the changes in the continuum of care with COVID-19 pandemic, we compared the medical visits and the dispensation of ART medications in our HIV-positive patients occurred between October 1st and November 30th, 2019 and March 1st and April 30th, 2020. We chose October and November 2019 for comparison, because they had similar amounts of festivities, same ARTs available and similar administrative management.

Statistical analysis
Categorical variables were described using frequencies and percentages, while continuous variables were described using mean, median, and interquartile range (IQR) values. Means for continuous variables were compared using independent group Student's t-tests when the data were normally distributed and the Mann-Whitney test when they were not. Proportions for categorical variables were compared using the χ2 test, although Fisher's exact test was used when the data were sparse.

Results
As of December 31st, 2019, a total of 3875 HIV-infected patients (mean age 51,4 ± 13 years old) were actively followed in our clinic. Of these, 28% were females and 18.8% non-Italian. Overall, 98.9% were on ART (n = 3834), of which 32,1% were assuming an ART regimen that included a protease inhibitor. Viral plasmatic load was < 20 cp/ml in 93% of patients and CD4 + > 350 cells/mm3 in 88.3% (data not shown).

New HIV diagnosis
During 2019, a total of 77 new HIV diagnosis were registered at our center, for a monthly mean of 6,4 new diagnosis. About 20% of all new diagnosis were in females. In the bimester of the COVID-19 pandemic we observed only ve patients with a rst HIV diagnosis (mean 2,5 new diagnosis per month). Of these, only one occurred in March, while the remaining four were made in April. Three out of ve new patients were females (60%), and only one of them was not-Italian (20%). AIDS-de ning conditions were present at the moment of HIV diagnosis in two patients (one had Kaposi sarcoma and the second one a non-Hodgkin lymphoma). Among the three remaining, two received HIV diagnosis during routine screenings for pregnancy and the last one in the context of a hospitalization for enteric intussusception. The median CD4 + T-cell count at diagnosis was 236 cell/mm3 (range 12-479 cell/mm3).

Number of medical visits at our HIV out-patients clinic
During the bimesters October-November 2019 and March-April 2020 a total of 1217 and 1162 patients had their visit scheduled at our out-patient HIV clinic, respectively. The overall view of the performed visits is shown in Fig. 2 and in Supplementary material 1. Thanks to the telemedicine program initiated at our clinic at the beginning of the COVID-19 pandemic, a total of 782 (67.3%) patients resulted eligible for this modality and performed their visit through this new useful tool in March and April 2020 (Fig. 2).

Dispensation of antiretroviral medications
We have found that the number of patients to whom ART was dispensed during March and April 2020 decreased of -23,1% in comparison with October-November 2019 ( Table 2). The highest decrease was observed in non-Italian compared to Italian (-27.4% vs -20.1%) and in females in comparison with males (-23.6% vs -20.5%). When we compared March and April separately with the mean of patients to whom ART was dispensed during the rst bimester, we observed a higher drop of drugs dispensation in March (-33.6%) with a trend to normalization in April (-12.6%).
Lastly, we explored potential differences among the number of drug packs dispensed in the two bimesters. We compared the number of drug packs dispensed in March-April 2020 and the mean of packs dispensed between June and December 2019. To this end, we chose to analyze the trends of the most used single tablet regimens (STRs) in our center, including an integrase inhibitor (lamivudine/abacavir/dolutegravir) and a protease inhibitor (PIs) (darunavir/cobicistat/tenofovir alafenamide/emtricitabine) (data not shown). Both STRs registered an average drop of -21.6% and − 16.6% respectively for the 2020 bimester in comparison with the second semester of 2019. *The percentage was calculated as the difference in the overall number of patients that collected ART between the different study periods.

Number of HIV-positive patients hospitalized
During 2019, a total of 92 HIV-positive patients were admitted in our ward (7,7 patients/month). Between March and April 2020, 25 HIV-infected patients were admitted to our Infectious Diseases clinic (12,5 patients/months). Among these, 12 patients were hospitalized because of SARS-CoV-2 infection (48%).
The great majority of patients reported fever at symptoms onset (n=9, 75%), dry cough (n=7, 58%) and dyspnea (n=3, 25%  is the rst step towards initiation into the HIV care continuum. Our clinic receives patients from a large province with more than 1 million inhabitants, where the incidence of HIV infection is estimated to be one of the highest in Italy, 5.8/100000 compared to 4.7/100000 in the whole country (7). From a public health perspective, the reduction of the occurrence of new cases is detrimental in order to reduce viral circulation. Promoting HIV testing with early diagnosis represents the essential entry point for both treatment and prevention efforts. Here we show as the COVID-19 pandemic had a negative impact on HIV screening programs in our province. Indeed, we observed a drop in the number of new HIV cases in comparison with the monthly mean of new HIV diagnosis occurred in the 2019 (2.5 vs 6.7 new HIV cases/month, p = 0.01). On March 8th, 2020 the Italian government declared the implementation of community containment measures in order to contain the diffusion of SARS-CoV-2 (8), which included social distancing, movement restrictions and quarantine for certain or suspected cases. Predictably, these measures drastically reduced the access to routine HIV testing (9,10). For a near future, it will be important to organize health services in order to guarantee a safe and continuous access to HIV testing, which is a fundamental health service and a detrimental step towards HIV elimination. Females and non-Italian patients resulted to be less adherent to follow-up visits during the pandemic period. This data needs to be further investigated and addressed.Anyway we believe that it is of fundamental importance to identify the more at-risk populations as auspicated by Hargreaves et al (11), in order to better address their needs and behaviors.
The second target (3), (90% of all people with HIV diagnosis should be maintained linked to HIV care) is based in offering and maintaining HIV care in all patients. Soon, hospital visits may be restricted again because of implementation of city lockdowns or tra c controls. From the beginning of March, we adopted the telemedicine tool with the purpose of avoiding the interruption of the continuum of HIV care. We performed structured phone interviews, where the physician together with collecting information about clinical status and ART adherence, also shared laboratory results and future appointments with HIVpositive patients. Telemedical consultations have helped to solve the otherwise unmanageable discrepancy between the #stayathome strategy and the need of a continuous medical assistance as also Ridgway et al experienced (12). Our preliminary experience shows that a large proportion of patients (67,3%) is eligible for a telemedical appointment. In the next years, we may think of this strategy for semesterly evaluation of stable patients. Moreover, this new tool may be helpful to avoid stigma related to be seen in an Infectious Diseases clinic, to decrease the loss of hours of work for medical visits and to reduce the illness perception. Nonetheless, we recommend that more complex cases keep on being considered for in-person medical evaluation and more often than every six months. Anyway, for a more comprehensive understanding of the e cacy of this approach and its impact on HIV care, we will need to evaluate HIV plasmatic suppression in the months to come.
The last target (3), (90% of HIV-positive patients with suppressed viral plasmatic loads) is based in assuring antiretroviral therapy to HIV-positive patients and in reinforcing adherence to ART. According to Italian regulations, HIV-positive patients need to collect their ART at the hospital pharmacy either personally or through a person authorized with a written approval form. Immediately after the beginning of the outbreak, many local charities began to offer home delivery services for ART medications through volunteers, in order to avoid patients exposure. At the beginning this service was not homogenous, relying on local associations or individual initiatives (e.g. municipal police, veterans, local charities, local Red Cross committee, etc..), while at the end of March, this service was coordinated on a national basis, and supported by Red Cross volunteers. In fact, we observed that although the mean decrease during the study period was − 23% compared to 2019, in March the decrease reached − 33,6%, when in April this trend tended to normalize (-12,6%). Once again, we will need to wait to better understand the consequences of this period on ART adherence and HIV replication.
Because many HIV-positive patients contacted us asking whether protease inhibitors would have been protective against SARS-CoV-2 infection (13-16), we analyzed the changes on the collection of STRs including protease inhibitors (darunavir/cobicistat/tenofovir alafenamide/emtricitabine) and STRs not including PIs (lamivudine/abacavir/dolutegravir) at our pharmacy. What we observed is that STRs including PIs had a smaller average drop than STR including integrase inhibitors (16,6% vs 21,6%, p < 0.05). Although we cannot assert whether patients on PIs have been more compliant to ART collection because of the possible effect of these drugs on COVID-19, we think that this data is interesting for it to be pointed out. Most of them were admitted with a moderate COVID-19 disease and had optimal outcomes, despite a higher average age and a higher proportion of patients with comorbidities (75%) in comparison with other cohorts (17,18).
Besides, all patients, except three, had CD4 + T-cells > 350/mm3, indicating no severe immune de ciency. As more data were coming out for the e cacy of PIs against SARS-CoV-2 in vivo (19), half of our patients maintained their usual ART and did not switch to a PI during the time of hospitalization. Of those hospitalized, only one patient was prescribed chronically with darunavir/cobicistat and was virosuppressed at admittance. Our observation, although very modest, may add evidence for a lack of a protective role of darunavir against SARS-CoV-2 (20), while more robust evidences are awaited (21). At last, our experience does not support the idea of an excess of morbidity and mortality among PLWH with viral suppression affected by COVID-19 pneumonia.
Moreover, we agree with the recent suggestion from Jones et al (22) of not modifying ART in order to attempt to treat SARS-CoV-2 infection.
Our study has important limitations. First, this is a retrospective study without follow-up, therefore it is not possible to ascertain the consequences and the e cacy of our strategies on the last 90 of our strategy (90% of all people receiving antiretroviral therapy with viral suppression). Moreover, the number of hospitalized HIV-positive patients may be small to offer a comprehensive understanding of COVID-19 disease in HIV-positive patients. About this, we may have "lost" some HIV/SARS-CoV-2 coinfections, due to eventual hospitalizations in hospitals other than ours. It must be said anyway, that HIV patients normally tend to be centralized and hospitalized in our ward.
Strength of this study is to have analyzed the main falls in the continuum of HIV care in our area during the emergency period. Now, we have learnt that telemedicine and home delivery services for ART medications are useful and powerful tools. Anyway, our data also emphasize the need to detect the populations at high risk for HIV care attrition (in our case, females and non-Italian patients) and to implement effective retention tools focused on these. We fell in maintaining high levels of routine HIV screening, which is crucial for early diagnosis and for reduction of viral circulation. Therefore, we learned that we have to keep on offering valid alternatives for HIV screening, where several health-care services and other community-based organizations were closed in this period due to the COVID-19 emergency. Moreover, in-person evaluations need to be maintained for those cases which are considered clinically more complicated or more at risk of loss.

Conclusions
Even though we cannot have de nitive data in such a short period, our preliminary impressions are that telemedicine can be a useful tool, which can make communications with some group of patients who are already engaged on the continuum of care available and easier. This strategy may be extremely usefull during pandemic times, but also beyond them and may be further implemented, for example through video calls. HIV testing services and strategies need to be supported and promoted even in critical periods, in the near future. Females and non-Italian patients are an especially at risk population for loss at follow-up, at least in the short period, so that special attention should be paid to these population in order to make them feel safe and to let them understand the fundamental importance of a continuous adherence to ART even during in emergency periods. Unfortunately, we may say, it seems like that we need to adapt our HIV care services to emergency settings. We hope that the strategy that we found will result effective in the near future, and that our experience Moreover, for this study we used the general authorization of the Italian Guarantor for the use of retrospective demographical and clinical data, which have been treated according to present laws.

Consent for publication
See above Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests   Proportion of patients that performed the scheduled medical visit in the rst bimester in comparison with the second bimester according to sex and nationality.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.