Determinants of compliance to influenza and COVID-19 vaccination in a cohort of solid organ transplant patients in Puglia, Southern Italy (2017–2022)

ABSTRACT

Influenza and Coronavirus Disease 2019 (COVID-19) vaccination are recommended in both solid organ transplant (SOT) candidates and recipients. In Puglia, Southern Italy, an active vaccination offer program has been activated targeting these patients. This study aims at investigating vaccination coverage (VC) for both vaccines in a SOT patients’ cohort, as well as at identifying the vaccination compliance determinant. This is a retrospective, population-based study. The study population consists of the SOT patients who accessed Bari’s “Policlinico” General Hospital during 2017–2022. Patients were contacted and, after providing their consent, asked their immunization status regarding influenza and COVID-19 and whether they had already undergone transplant or were waiting to do so. Regression models were fitted to investigate the determinants of VCs for influenza vaccination (2021/22 and 2022/23 seasons) and for COVID-19 vaccination (three-dose base cycle, first and second booster doses). Three-hundred and ten SOT patients were identified; 85.2% (264/310) had already undergone SOT. VCs were suboptimal, especially for constant yearly influenza vaccination (17.7%) and COVID-19 vaccination’s second booster (1.94%). Logistic regression highlighted that influenza VCs are higher for SOT recipients than SOT candidates, as well as for older patients, although when considering both vaccination seasons only age significantly impact the vaccination uptake. Older age was the only influential variable for COVID-19 VC. VCs for SOT patients seem to be unsatisfying. Stronger interventions are required.

KEYWORDS:

• Hesitancy
• vaccination coverage
• safety
• high risk group

Background

Solid organ transplant (SOT) is associated with an increased risk of infections stemming from both the immune suppression caused by anti-rejection therapy and the baseline health conditions of transplanted subjects. Infections are also characterized by a higher morbidity and mortality in SOT patients. An especially high risk has been observed for viral respiratory infections, whose widespread circulation and ease of transmission makes SOT patients particularly vulnerable to this kind of infectious disease.^1–3

Seasonal influenza and Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection are currently two of the main viral respiratory diseases who may cause complications and death in SOT patients. They may be acquired in both community and healthcare settings and represent a threat to these subjects’ survival and quality of life. Further, serious infections in this vulnerable population may hinder national healthcare systems’ activities due to the increased workload on healthcare facilities.^4–7

Vaccination represents the main tool for the prevention of influenza and Coronavirus Disease 2019 (COVID-19) in general population and immunocompromised patients. It is essential to constantly monitor and update the safety and efficacy profile of the vaccine product in high-risk groups through specific post-marketing surveillance programs.^8–10 In fact, immune response to vaccination is generally lower in these patients than in the general population, thus requiring more doses or specific products in order to maintain an acceptable level of protection.^11–13

Italian Vaccination Schedule for Life recommends yearly influenza vaccination for all SOT recipients as well as for SOT candidates.¹⁴ In Italy, these patients were identified as priority targets for COVID-19 vaccination right after healthcare workers, and have been actively summoned for vaccination since March 2021. Subsequently, a note from the Italian Ministry of Health recommended additional COVID-19 vaccine doses for these subjects starting from five years of age, considering the higher risk of infection and hospitalization in patients with chronic diseases and/or immune compromission. Yearly booster doses have also been provided ever since for all subjects, but with special recommendation for immunocompromised patients.^15–18

As of now, data regarding COVID-19 vaccination coverage of SOT candidates and SOT recipients are not gathered routinely in Europe. As far as influenza vaccination is concerned, a few studies have been carried out investigating vaccination coverage in SOT patients, generally highlighting suboptimal values.¹⁹ In Italy, a countrywide investigation of influenza vaccination coverage in subjects suffering from chronic diseases is carried out periodically within the Progress of Healthcare Public Bodies for Health in Italy (PASSI surveillance project). The surveillance for years 2020–2021 estimated vaccination coverage in subjects with kidney failure and chronic liver disease at 34.2% and 18.3%, respectively, highlighting an alarming situation.²⁰

Our study evaluates influenza and COVID-19 vaccination coverage in a cohort of SOT patients/candidates in Puglia, Southern Italy. In Puglia, dedicated routes for vaccination offer have been activated for all SOT patients thanks to an agreement between the Regional Transplant Center and Bari’s “Policlinico” General Hospital. Following the first medical contact, these patients’ information is transmitted to the hospital’s vaccination clinic in order for them to be contacted and actively offered vaccination. Our main goal is to identify potential determinants for these subject’s compliance to vaccination recommendations. A subsidiary objective is to evaluate the effectiveness of Apulian organizational model for vaccination offer to highly vulnerable patients.

Materials and methods

This is a retrospective, population-based study. Our study population is represented by SOT candidates and SOT recipients who attended the “Policlinico” General Hospital of Bari, South-East of Italy, at least once for either hospitalization or ambulatorial procedures from January 1, 2017, to December 31, 2022. These patients’ names and cellphone numbers were obtained from the wards which provided them with these services.

The data extraction phase started on March 1, 2023, and ended on April 30, 2023. All subjects were contacted via phone call and informed about this study’s purpose and methodology. In case of minors, a legal guardian was contacted in their stead. Informed consent to participate to the study was therefore required.

Participants were therefore asked to provide their clinical documentation regarding the SOT they either had undergone or were candidate to receive. Other collected data included biological sex, date of birth and influenza and COVID-19 vaccination history. As far as influenza vaccination is concerned, only 2021/22 and 2022/23 seasons were taken into consideration. All data were anonymized during the database building phase. For context, influenza vaccination is offered yearly from October to December, while the COVID-19 vaccination has been offered since March 2021, with the third dose being administered during September and October 2021. The first and second booster dose have been offered in March/April and October/December 2022, respectively.

The “age” variable was categorized using the following age groups: 0–18 years, 19–45 years, 46–64 years and over 65 years. Qualitative variables were expressed as percentages (proportion). Discrete quantitative variables were expressed as median (interquartile range, IQR). Vaccination coverages (VCs) for the considered vaccines were calculated as follows:

$VC=\frac{vaccinationsubjects}{totalnumberofsubjectseligibleforvaccination}\times 100\mathrm{\%}$

Mean ages of the male and female subgroups were confronted via t-Student test for independent samples, since the variable distribution in both subgroups was proven to be normal. VCs in different population subgroups were confronted via Chi-squared test. A logistic regression model was fitted to verify whether the subjects’ sex, age group and SOT status at the time of vaccination (candidate or recipient) significantly impacted on VCs. Finally, another logistic regression model was fitted to study the effect of sex, age group, SOT status and influenza vaccination status on COVID-19 VC. A two-sided p-value <0.05 was considered as an indicator of statistical significance.

Data were digitalized via software Microsoft Excel®. All statistical analyses were conducted via software StataMP®. The study was carried out in accordance with the Declaration of Helsinki; since no biological samples were gathered and data was strictly anonymized, no ethical committee approval was necessary.

Results

During the study’s time frame, 310 SOT recipients/candidates were identified who had undergone hospitalization and/or ambulatory procedures in Bari’s “Policlinico” General Hospital at least once from 2017 to 2022. Bari’s General Hospital is the largest healthcare facility in Southern Italy, representing one of the main transplant centers of this geographical area.

Following phone contact, all subjects agreed to take part to the study. Out of all enrolled patients, 65.5% were males (203/310) and 34.5% were females (107/310), and 85.2% of them (264/310) had already undergone SOT by March 1, 2023.

The median age was 56 years (44–64), and no significant difference was identified in mean ages of female and male subjects (t: 0.92; p-value: .36). Following categorization, 3.5% of the subjects (11/310) were comprised in the 0–18 age group, 22.6% (70/310) in the 19–45 age group, 52.3% (162/310) in the 46–65 age group and 21.6% (67/310) were over 65 years of age. Since one patient was only 4, they were excluded from COVID-19 VCs calculation.

Influenza VC: 2021/22 influenza season

Only 39.4% (99/310) of the study population received their influenza shot during the 2021/22 influenza season. No significant difference was found in VCs of males and females (Pearson chi2: 0.31; p-value: 0.58), while a significant difference was identified between SOT recipients and candidates (Pearson chi2: 6.94; p-value <.05). Different age groups were also characterized by different VCs (Pearson chi2: 16.91; p-value <.05) (Table 1).

Table 1. Influenza vaccination coverage of patients by sex, age class and solid organ transplant status.

		N.	Vaccinated (%, 2021/2022 influenza season)	Vaccinated (%, 2022/2023 influenza season)	Vaccinated (%, yearly dose received in both 2021/2022 and 2022/2023 influenza season)
Sex	Male	203	67 (33.0%)	67 (33.0%)	41 (20.2%)
	Female	107	32 (29.9%)	31 (29.0%)	14 (13.1%)
Solid organ transplant status	Recipient	264	92 (34.8%)	88 (33.3%)	52 (19.7%)
	Candidate	46	7 (15.2%)	10 (21.7%)	3 (6.5%)
Age class	0–18 years	11	3 (27.3%)	2 (18.2%)	1 (9.1%)
	19–45 years	70	9 (12.9%)	11 (15.7%)	3 (4.3%)
	46–64 years	162	58 (35.8%)	54 (33.3%)	29 (17.9%)
	≥65 years	67	29 (43.3%)	31 (46.3%)	22 (32.8%)

The fitted logistic regression model confirmed a significant effect of SOT status and age group, with higher VC for SOT recipients than candidates (OR: 2.98; 95% Confidence Interval, 95 CI: 1.26–7.05; p-value <.05) and for older subjects (OR: 1.84; 95 CI: 1.29–2.61; p-value <.05). No significant influence of sex was identified (OR for males: 0.97; 95 CI: 0.57–1.65; p-value: .92).

Influenza VC: 2022/23 influenza season

During the 2022/23 influenza season, VC in the study population further lowered reaching 31.6% (98/310). Once again, females and males did not present significant differences in terms of VC (Pearson chi2: 0.53; p-value: .47), whereas different age classes had significantly different coverages (Pearson chi2: 15.98; p-value <.05). However, SOT status appeared not to be relevant for this season’s influenza VC (Pearson chi2: 2.44; p-value: .12) (Table 1).

Logistic regression confirmed these results, highlighting a relevant increase in VC for older age groups (OR: 1.95; 95 CI: 1.37–2.76; p-value <.001). Sex (OR for males: 0.98; 95 CI: 0.53–1.52; p-value: .68) and SOT status (OR for recipients: 1.74; 95 CI: 0.81–3.76; p-value: .16) were not showed to have any significant impact on VC.

Influenza VC: both influenza seasons

Only 17.7% of all subjects (55/310) received influenza vaccination during both study seasons. Sex was not related to significant VC differences (Pearson chi2: 2.43; p-value: .12). Both SOT status (Pearson chi2: 4.66; p-value <.05) and age groups (Pearson chi2: 19.71; p-value <.05) were associated to a significant difference in VC (Table 1).

The logistic regression model partially confirmed observed findings. In fact, older age was the only independent variable significantly increasing VC (OR: 2.61; 95 CI: 1.64–4.16; p-value <.05). On the other hand, neither having already received the transplant (OR: 3.41; 95 CI: 0.99–11.77; p-value: .052) nor the male sex (OR: 0.66; 95 CI: 0.33–1.30; p-value: .23) had a significant effect on VC.

COVID-19 VC: primary cycle

VC for the COVID-19 vaccine’s three-dose primary cycle was 81.2% (251/309). No significant differences were highlighted between males and females (Pearson chi2: 2.45; p-value: .12). Age groups were characterized by a statistically significant difference in VC (Pearson chi2: 45.95; p-value <.001) and a significant difference was also observed for VCs in SOT recipients and candidates (Pearson chi2: 4.83; p-value <.05) (Table 2).

Table 2. COVID-19 vaccination coverage of patients by sex, age class and solid organ transplant status.

		N.	Vaccinated (%, primary three-dose cycle)	Vaccinated (%, first booster dose)	Vaccinated (%, second booster dose)
Sex	Male	203	170 (83.7%)	75 (36.9%)	Vaccinated (%)
	Female	107	81 (76.4%)	24 (22.6%)	6 (3.0%)
Solid organ transplant status	Recipient	264	219 (83.3%)	88 (33.5%)	–
	Candidate	46	32 (69.6%)	11 (23.9%)	5 (1.9%)
Age class	0–18 years	11	1 (10.0%)	–	1 (2.2%)
	19–45 years	70	49 (7.0%)	12 (17.1%)	–
	46–64 years	162	144 (88.9%)	55 (33.9%)	1 (1.4%)
	≥65 years	67	57 (85.1%)	32 (47.8%)	2 (1.2%)

When entered into a logistic regression model, VC proved to be significantly dependent on age groups, with older groups having higher coverage (OR: 2.53; 95 CI: 1.68–3.79; p-value <.001). No significant effect was observed from sex (OR: 0.68; 95 CI: 0.36–1.26; p-value: 0.22). SOT status did not appear to be influent on VC either (OR: 1.95; 95 CI: 0.92–4.15; p-value:.08).

COVID-19 VC: first booster dose

A significant drop in VC for the COVID-19 vaccine’s fourth dose, as only 32.0% of subjects (99/309) received it. Interestingly, a significant difference in terms of VC was observed between males and females (Pearson chi2: 6.54; p-value <.05) and between the different age groups (Pearson chi2: 19.73; p-value <.001), while different SOT statuses had no significant differences as far as coverage is concerned (Pearson chi2: 1.64; p-value: .20) (Table 2).

Logistic regression proved an increased odd of vaccination for the male sex (OR for females: 0.51; 95 CI: 0.29–0.89; p-value <.05) and for older subjects (OR: 2.21; 95 CI: 1.53–3.18; p-value <.001). SOT status was confirmed to be uninfluential (OR: 1.41; 95 CI: 0.66–3.02; p-value: .38).

COVID-19 VC: second booster dose

Only 1.94% subjects were vaccinated with a fifth dose of COVID-19 vaccine (6/309). Furthermore, no significant differences in VC were highlighted between males and females (Pearson chi2: 3.19; p-value: .07), different age groups (Pearson chi2: 2.98; p-value: .39) and transplantation status whatsoever (Pearson chi2: 0.01; p-value: .90). Since no significant differences were observed, no regression model was fitted (Table 2).

Interaction between influenza and COVID-19 vaccination coverage

The primary cycle COVID-19 VC proved to be significantly different between subjects who had and had not been vaccinated against influenza during each of the considered influenza seasons (Pearson chi2, 2021/22 influenza season: 20.73; p-value <.001; Pearson chi2, 2022/23 influenza season: 10.59; p-value <.05), as well as between SOT patients who had and had not undergone influenza vaccination during both campaigns (Pearson chi2: 12.61; p-value <.001) (Table 3).

Table 3. COVID-19 vaccination coverage of patients by influenza vaccination status (primary three-dose cycle).

		COVID-19 vaccination status
		Vaccinated	Unvaccinated	Total
Influenza status (2021/22 season)	Vaccinated	95 (96.0%)	4 (4.0%)	99
	Unvaccinated	156 (74.3%)	54 (25.7%)	210
Influenza status (2022/23 season)	Vaccinated	90 (91.8%)	8 (8.2%)	98
	Unvaccinated	161 (76.3%)	50 (23.7%)	211
Influenza status (both seasons)	Vaccinated	54 (98.2%)	1 (1.8%)	55
	Unvaccinated	197 (77.6%)	57 (22.4%)	254

However, logistic regression with sex, age group and SOT status as covariables highlighted a significant association only for the 2021/22 influenza season vaccination status, which was associated with a higher COVID-19 VC (OR: 5.09; 95 CI: 1.46–17.72; p-value <.05). Influenza vaccination uptake for the 2022/23 season was uninfluential (OR: 1.76; 95 CI: 0.71–4.35; p-value: .22), as well as the vaccine uptake for both influenza seasons (OR: 1.51; 95 CI: 0.12–18.25; p-value: .75).

As far as the first booster dose of COVID-19 vaccine is concerned, a significant difference was observed in terms of VC between subjects who had and had not undergone influenza vaccination during the 2021/22 (Pearson chi2: 33.89; p-value <.001) and 2022/23 influenza season (Pearson chi2: 41.54; p-value <.001). Subjects with and without two influenza vaccine doses during the study period also had different VC for the first COVID-19 booster (Pearson chi2: 50.87; p-value <.001) (Table 4).

Table 4. COVID-19 vaccination coverage of patients by influenza vaccination status (first booster dose).

		COVID-19 vaccination status
		Vaccinated	Unvaccinated	Total
Influenza status (2021/22 season)	Vaccinated	54 (54.5%)	45 (45.4%)	99
	Unvaccinated	45 (21.4%)	165 (78.6%)	210
Influenza status (2022/23 season)	Vaccinated	56 (57.1%)	42 (42.9%)	98
	Unvaccinated	43 (20.4%)	168 (79.6%)	211
Influenza status (both seasons)	Vaccinated	40 (72.7%)	15 (27.3%)	55
	Unvaccinated	59 (23.2%)	195 (76.8%)	254

When tested in a logistic regression model with sex, age group and SOT status, influenza vaccination during each season proved to significantly increase the first booster dose VC (2021/22 OR: 2.25; 95 CI: 1.03–4.92; p-value <.05; 2022/23 OR: 2.84; 95 CI: 1.33–6.06; p-value <.05). The influenza vaccination two-season uptake, on the other hand, was not proven to significantly impact the booster dose’s uptake (OR: 1.61; 95 CI: 0.50–5.20; p-value: .43).

The second booster dose was not tested in regard to influenza vaccination status due to the low numerosity of vaccinated subjects.

Discussion

The SOT patient cohort we recruited from Bari’s “Policlinico” General Hospital’s attendants showed alarmingly low VC for both influenza and COVID-19 vaccination. Only a small percentage of the study subjects (17.7%, 55/310) complied steadily to influenza vaccination campaigns over the two considered seasons, and even fewer patients (1.94%, 6/309) received the second COVID-19 vaccine’s booster dose. The primary cycle of the latter, on the contrary, was accepted by a larger number of subjects (81.2%, 251/309).

Although the primary COVID-19 cycle’s VC was not significantly influenced by influenza vaccination uptake except from the 2021/22 season, the first booster dose’s uptake significantly increased in subjects who had received influenza vaccination during the 2021/22 and 2022/23 seasons. Interpretation of these results is convoluted, as it revolves around different factors. First of all, in 2021, the first three doses of the COVID-19 vaccination were made mandatory by the Italian Government in order to access various services.²¹ Secondly, the strong expectation for COVID-19 vaccines might have distorted the general public’s perception of this vaccination, separating it from commonly administered vaccines.²²

On the other hand, the first booster dose was first offered in 2022, after the dismission of the COVID-19 vaccination mandate in Italy. It is legitimate to hypothesize that subjects who accepted to receive this dose had greater health literacy and/or a generally better attitude toward vaccination, health literacy being associated with a higher acceptance of seasonal influenza vaccination.^23,24 It is also interesting to highlight that a French study observed better compliance to influenza vaccination for subjects who have stabler contacts with their physician.²³ However, none of these studies concerned solely SOT patients, rendering these explanations only partial; more investigations are required in order to get a better understanding of vaccination acceptance in highly vulnerable subjects.

Vaccination acceptance in SOT candidates and recipients has often been proven to be unsatisfactory. A 2021 Danish cross-sectional survey reported a 41.8–53.2% influenza vaccine uptake in kidney transplant recipients and waiting list patients combined, with the latter having a higher acceptance rate than the former.²⁵ A Belgian study from 2020 also highlighted suboptimal VC for influenza in SOT recipients, with only 52.9% of these subjects vaccinated against influenza during the previous influenza season.²⁶

On the other hand, only a few studies exist regarding COVID-19 VCs in SOT patients. A 2022 Indian multicentered study identified significant gaps in coverage among different hospitals and transplant centers: 8.5% of the enrolled centers reported that less than 25.0% of their patients had been fully vaccinated before undergoing transplant, while 70.2% of them reported a VC above 50.0% in their own population.²⁷ More recently, a 2023 Chinese survey reported extremely low three-dose VC in liver transplant recipient children (0.9%), despite higher acceptance rates being reported by these subjects’ guardians.²⁸

Our findings are therefore in line with current knowledge, confirming that vaccination efforts are still far from reaching their optimal results in a highly vulnerable population such as SOT patients. The question should be answered, however, whether the low VCs we observed should be considered a consequence of the patients’ vaccine hesitancy or of their physicians’ hesitancy to recommend vaccination.^29,30 Whatever the case, the fall of VCs over time is concerning, and its underlying causes are to be investigated further.

Significant progresses have been recorded over the last decades in terms of attitude of healthcare workers toward recommendation of vaccines to their patients.^31–33 Nevertheless, a multicentered survey carried out on 141 transplant facilities in the United States highlighted that 60.7% of these centers currently do not require COVID-19 vaccination before transplant, often due to legal liability concerns.³⁴ Despite effectiveness in reducing infections over time and preventing hospitalization and death, a small proportion of healthcare workers are hesitant toward SARS-CoV-2 vaccination or straight up refuse it, mainly due to safety concerns; this lack of trust is expected to also influence these workers’ attitude toward recommending vaccination to their vulnerable patients.^35–39

As far as the patients are concerned, various studies have demonstrated a significant degree of hesitation among SOT recipients/candidates, especially for the COVID-19 vaccination. A Chinese paper’s study population had a 66.6% prevalence of vaccine hesitancy, generally caused by fear of adverse events.⁴⁰ In a similar fashion, a study carried out in America in 2020 showed that almost half of the enrolled kidney transplant recipients was hesitant to undergo COVID-19 vaccination, the main reasons being safety concerns and fear of incompatibility with organ transplant.⁴¹ Rejection and hesitancy in SOT recipients were also observed for influenza vaccination, with a small study in Sicily (South of Italy) identifying significant fear of vaccine-related adverse events as one of the main drivers for vaccination refusal.⁴² On the other hand, the systematic use of causality assessment showed that vaccines are safe in SOT recipients and not related to rejection or other major adverse events.^43–46 Considering WHO and Italian Health Minister recommendation it is mportant to estimate safety profile and the incidence of SARS-CoV-2 breakthrough infections in subjects receiving the SARS-CoV-2 booster and influenza vaccines simultaneously.^47–50

Our study presents a few limits, the main one being the low numerosity of our sample. However, since it is a population-based study, our findings accurately reflect the real-world VC for SOT patients in a large Italian hospital. An additional weakness is the lack of information regarding our patients’ socio-economic status and education level, which has been observed to significantly impact vaccination rates in previous studies.^51–53 This would be an interesting development for future research, which could also help tackle social disparities via the personalization of vaccination offer according to socio-economic factors. Finally, it was impossible for us to determine how many SOT patients currently live in Puglia. Our data might therefore be an overestimate of real-world VCs, as we only covered subjects who did attend follow-up before or after transplantation. Those who are not adequately followed may have even lower VCs, thus accounting for an even worse general scenario.

On the other hand, we had an unexpectedly high participation rate thanks to the long-lasting physician–patient relationship our transplant center’s staff were able to build with their patients. We gathered data from one of the main transplant centers in Italy, reaching one of the largest SOT recipient populations available in the country. Finally, we were able to obtain a full vaccination history for all patients thanks to the availability of a detailed online database.

Conclusion

Vaccination coverage in SOT recipients and candidates is still unacceptably low, both for the influenza and COVID-19 vaccination. New policies should be taken into consideration to improve vaccine acceptance among these patients, as well as to make vaccination necessary in order to undergo transplantation. Finally, tailor-made communication efforts should be provided by healthcare professionals during these patients’ pre-transplant phase.

Our organizational model, although relatively well performing during the first pandemic phase, rapidly underperformed over time. As already stated, this might have been associated with the lower risk perception among both the patients and the medical staff, but the return to normalcy might also have induced physician to go back to their pre-pandemic routine, thus neglecting vaccination. As a future development, a more continuous and hands-on approach could be enforced, with a closer relationship between transplantation centers and vaccination clinics. Vaccination-specialized facilities could be activated in a wider network over the Apulian territory in order to facilitate the patients’ access, especially for those who cannot move by themselves over long distances.

Improving VCs is fundamental in order to contain costs, reduce infectious disease-related mortality in high-risk subgroups and avoid medico-legal issues in case of complications and death. Both national and regional level institutions and hospital directions should therefore strive to thoroughly inform and follow up these patients in order to grant them the best possible medical care, including preventive measures.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.