Pneumococcal vaccination coverage in individuals (16–59 years) with a newly diagnosed risk condition in Germany

Our results indicate a very low pneumococcal VCR in individuals aged 16–59 years with a newly diagnosed risk condition in Germany. These findings are in alignment with other studies in Germany [10‐12].

Theidel et al. estimated the pneumococcal VCR for the one-year period July 1, 2008 to June 30, 2009 in individuals aged 18–59 in Germany, using data from the Deutsche BKK health insurance database [10]. The VCR was estimated as 1.32% in individuals at medium risk for PD (diagnosed with a chronic cardiovascular, respiratory, metabolic, renal or neurological disorder) and 0.90% in individuals at high-risk for PD (diagnosed with an immunocompromising condition). The VCR was reported as 0.23% in individuals aged 18–59 not-at-risk of PD (i.e. not recommended for pneumococcal vaccination according to STIKO). These percentages differ from the results seen in the present study. However, the studies are difficult to compare due to differences in methodology, including the criteria for selection of the study population and risk group definitions. In contrast to the study by Theidel et al. the present study found higher VCRs in individuals with high-risk vs. at-risk conditions, as would be expected – likely explained by the different risk definitions (coding) and study inclusion criteria applied by Theidel et al., in addition to their very short study period of 1 year. Interestingly, Theidel et al. reported a higher (but still relatively low) VCR of 4.45% in adults aged 60 and over, a population who are eligible for vaccination based on older age.

Schmedt et al. investigated pneumococcal VCRs from 2014 to 2016 in adults aged 16–59 years with a newly diagnosed high-risk (immunocompromising) condition in Germany using the InGef database [11]. The authors reported a VCR of 1.6% in women, and 2.3% in men; broadly in alignment with results from the present study. Median time to vaccination was estimated at 332.5 (IQR 142–528) days (within the first two years of diagnosis) [11]—in agreement with the median time to vaccination of 369.5 (IQR 196.3–758.8) days in the present study (high-risk group). In a follow-up study to Schmedt et al., Sprenger et al. reported pneumococcal VCR within two years of a diagnosis of an incident high-risk condition, for individuals diagnosed in 2012/2013 (cohort A) versus those diagnosed in 2015–2017 (cohort B) [12]. In agreement with results reported in the present study, VCRs remained very low across both 2-year follow-up periods: estimated at 2.0% and 2.5% for cohort A and cohort B, respectively.

Increasing the VCR in individuals aged 16–59 years with newly diagnosed risk condition(s) is crucial, as there is an anticipated rise in chronic conditions diagnosed worldwide in this age group, associated with an increased risk of PD and hospitalization due to PD [5, 6, 16, 17]. A possible explanation for the low VCR and delayed time to vaccination observed in Germany could be due to the uncertainty among general practitioners and other medical professionals as to who is responsible for administering vaccinations. Schmedt et al. [11], reported that 93.2% of all vaccinations were administered by general practitioners, with similar results reported more recently by Sprenger et al. [12]. However, patients with a newly diagnosed risk condition are likely to be treated by specialists such as pneumologists, oncologists or endocrinologists and not only by general practitioners. These specialists are not all reimbursed for PD vaccinations or receive training on the STIKO vaccination guidelines and may therefore refrain from recommending vaccination. Therefore, there may be a need for raised awareness of PD-vaccination recommendations across the medical specialties who diagnose and treat patients with PD-associated risk conditions.

Furthermore, although STIKO issues general recommendations for all German federal states, there is no national immunization plan in place for pneumococcal vaccination in Germany. Some federal states even have their own vaccine committees, which can issue different recommendations to STIKO. The UK, which has a clear national immunization schedule and guidance for healthcare professionals [18] and centralized vaccine procurement, has reported much higher VCRs in at-risk populations not otherwise eligible for pneumococcal vaccination based on age. A recent UK study using data from the Clinical Practice Research Datalink reported VCRs of 13.6% in the first year of follow-up, rising to 32.0% after 4 years of follow-up between 2011 and 2015, for individuals at-risk or high-risk for PD due to underlying medical conditions [19].

Neufiend et al. recently highlighted decentralized vaccine procurement as a key contextual barrier to vaccine provision in Germany, as it makes distribution and reimbursement more challenging [20]. Interestingly, the authors reported that encouraging or incentivizing physicians to get vaccinated themselves may be an alternative avenue to improve the low adult vaccination rates in Germany. In addition to increasing institutional trust in STIKO. A survey conducted in approximately 5,000 practices in Germany has highlighted the need for automated recall/reminder-systems, implementation of routine, regular vaccination counselling and education of physician assistants to help reduce the risk of neglecting to advise patients on vaccinations [21, 20]

Our study reports a higher likelihood of pneumococcal vaccination in those who have received the influenza vaccination. Indeed, many individuals aged 16–59 for whom pneumococcal vaccination is indicated are also likely to be eligible for influenza vaccination. These two vaccines can be administered in the same session [22], and may provide a solution to increasing the pneumococcal VCR. The impact of the new coronavirus vaccination program in 2021 [23] is unknown, but may present another opportunity to increase the pneumococcal VCR.

There is currently no recommended minimum VCR for pneumococcal vaccination in Germany or internationally. Yet it is evident that there is a strong need for health policy makers to implement further clinical practice guidelines at a national level. Training of physicians and prompting at the time of diagnosis of a risk condition may also improve the VCR considerably in these vulnerable patient groups.

The main limitation of this study is the potential selection bias of the study population. The InGef database is comprised only of individuals with state health insurance and may not include all sociodemographic strata of the German population, which may have led to a biased estimation of VCR. Furthermore, due to the structure of the SHIs that provide data for the InGef research database, the proportion of insurees in the regions of East/ West Germany differs slightly from the German population; a smaller proportion of insurees residing in the eastern part of Germany for the InGef database (16.5%, versus 19.5% in the general population) [24]. Nevertheless, previous studies have demonstrated that the InGef database is representative of the German general population for age and sex, and when studying health outcomes, morbidity, and drug usage (14).

Another limitation was the inability to distinguish the different types of pneumococcal vaccines administered (e.g., PCV13 vs. PPSV23). Information on manufacturer brand is not available in the German research databases [25]. We therefore could not assess whether those individuals with an indication for sequential vaccination (based on STIKO recommendations) received PCV13 or PPSV23 or both. Further studies would benefit from using different data sources, in order to allow for the distinction between different types of vaccines (PCV13 vs. PPSV23) and to investigate delivery of sequential booster vaccinations, where indicated by STIKO. In addition, it would be useful to also assess the proportion of vaccines delivered by general practitioners versus other healthcare professionals/specialists to further understand where the pneumococcal VCR could be improved. However, Sprenger and colleagues reported that > 90% of pneumococcal vaccinations administered to immunocompromised (high-risk) patients were done so by a general practitioner, and rarely by other specialists [12].

Finally, there may be misclassification bias due to coding inaccuracies since medical conditions were identified based on administrative records. For example, individuals with history of risk conditions before the pre-period may have been classified as newly diagnosed. With only a two year look back before the index date it is possible that individuals with previous pneumococcal vaccination were misclassified as unvaccinated. The effect of this misclassification would be to underestimate pneumococcal VCRs in this population. It is possible that individuals with previously stable chronic conditions, such as asthma or diet-controlled diabetes, were misclassified as not at-risk. The effect of this misclassification would be to underestimate the eligible population for pneumococcal vaccination. However, it would be unlikely to bias the VCR in the selected population.