Range of invasive meningococcal disease sequelae and health economic application – a systematic and clinical review

Invasive meningococcal disease (IMD) is a life-threatening infectious disease, caused by the gram-negative bacterium Neisseria meningitidis (N. meningitidis) [1]. Disease incidence is highest in infants and young children, followed by adolescents i.e., 8.2 microbiologically-confirmed cases per 100,000 population in infants under 1 year of age, 2.5 in children aged 1–4 years, and 1.0 in adolescents aged 15–24 years across the European Union/European Economic Area in 2016 [2‐4]. Overall incidence rates are considerably lower in the United States (US) (overall 0.13/100,000 in 2017), yet they follow similar age-related patterns [5].

IMD has a fluctuating epidemiology with variable N. meningitidis serogroup distribution worldwide [6, 7]. Six N. meningitidis serogroups (A, B, C, W, X and Y) are responsible for most IMD cases worldwide [5]. Serogroup B is currently the most prevalent in Europe and the US [2, 3], but there are increasing reports of serogroup W cases from Europe, South America and Australia [7]. IMD can be prevented with available vaccines targeting serogroups C, ACWY or B. The national immunisation programs of different countries, however, include different vaccines for IMD prevention. Following the introduction of meningococcal conjugate C vaccines in 1999 (and later quadrivalent meningococcal ACWY vaccines), initially in the United Kingdom (UK), the incidence of IMD due to serogroup C declined in both vaccinated and unvaccinated populations, due to herd immunity [8]. The adoption of meningococcal B vaccine (4CMenB) into the national immunisation programs in the UK and regions of Italy has led to further important reductions in IMD incidence caused by serogroup B [9, 10] and W IMD (using modelling to estimate the direct impact of 4CMenB, based on real-world serogroup W data, pre- and post-vaccination) [11]. A retrospective cohort study in the UK of IMD cases diagnosed between 2008 and 2017 shows a decrease in IMD in young children 0–4 years of age and in adolescents reflecting the introduction of 4CMenB infant vaccination in 2015 and adolescent quadrivalent ACWY vaccination in 2016 [12].

Common clinical presentations of IMD include meningitis and septicaemia, which can lead to high case fatalities in the acute phase, as well as permanent, and sometimes very severe, sequelae in IMD survivors [13]. The long-term sequelae of IMD significantly impact the health of patients as well as their families and social contacts in general. However, the IMD burden beyond the individual patient is poorly defined [14].

It is important to comprehensively assess the impact of the disease burden when evaluating interventions to prevent or treat IMD [12, 13]. Several gaps in the literature have been identified [14] e.g., studies in IMD survivors may not always capture sequelae that take time to develop, due to insufficient follow-up. Because IMD is uncommon and can produce a wide range of sequelae, observational studies (OS) typically contain small numbers of patients and report different sequelae, and few are case-control studies. Economic evaluation (EE) in IMD is often used to inform policy decisions regarding new preventative or therapeutic interventions. It is, therefore, crucial to systematically identify, define, and evaluate all potential IMD sequelae to understand the full disease burden and to define the clinical burden of interest to policy decision-makers.

The objectives of this study were 1) to comprehensively map the broad range of IMD sequelae with respect to manifestation, severity and duration, via systematic literature review (SLR), and 2) to provide a systematic approach to identify the clinical impact of sequalae of relevance to EE to inform policy decision-making. A video summary linked to this article can be found on Figshare: https://​doi.​org/​10.​6084/​m9.​figshare.​19753840.

This study comprehensively analysed observational and economic studies from 2001 to 2020, to determine the clinically-relevant burden of IMD sequelae. A particular focus was on the sequelae included in EE. Using a systematic approach, 44 sequelae of varying severity were included in the comprehensive map, of which 18 were classed as important for EE. The physical/neurological and psychological/behavioural sequelae were caused by any serogroup, and some were permanent and disabling.

With the exception of 6/66 OS and 4/34 EE that reported over 10 IMD sequelae, most other studies reported a narrow range of sequelae, with on average only one psychological/behavioural sequela (Fig. 3). It is important to note that physical/neurological sequelae are often apparent e.g., amputation, skin scarring or hearing loss, and are, thus, likely to be reported. In contrast, certain psychological/behavioural sequelae are more insidious, may take time to become apparent, or may not be attributed to IMD, and thus may evade reporting. In a recent retrospective case-control study of IMD in the UK, psychological sequelae were reported after a median of 15.5 months (up to 36.2 months in infants) while neurological sequelae were reported after a median of 8.5 months and physical sequelae, a median of 1 month [12]. As such, the risk of under-reporting and under-estimation of IMD burden due to psychological/behavioural sequelae is high. The design of OS intended to capture the full extent of psychological/behavioural sequelae of variable severity and duration is demanding, since extended observation periods are required, as are sophisticated designs including blinded investigators, adequate control groups, tools to determine coping mechanisms [32], and so forth.

EEs tend to focus on quality-adjusted life-years and cost impact in cost-effectiveness analysis, therefore, sequelae that have an important adverse effect on health-related quality of life (HRQOL) or survival, and those with long-term or high costs, are particularly relevant for EE. EEs with a more comprehensive approach e.g., those including a wide range of sequelae with their associated impact on HRQOL of patients and caregivers as well as long-term costs [22‐24] are better able to depict the cost-effectiveness of IMD interventions, an essential factor when used to guide policy decisions [33]. By contrast, EEs that include a narrow range of sequelae may underestimate the burden of IMD and its health and economic impact, and thus fail to grasp the cost-effectiveness of IMD interventions, as demonstrated in Beck et al. (2021) [15]. A recent systematic review of economic evaluations in depression has shown that changing from a narrow payer perspective to a broader societal perspective (e.g., considering the burden on caregivers as well as patients) can change the outcome of the analysis and thus affect policy decisions [34]. Systematic reviews are better suited to capture the burden of disease and thus to be used in EEs, yet they rely on the quality of included studies.

The evaluation of study findings from a clinical and real-world perspective highlighted several gaps and areas for future research. Firstly, regarding the list of possible sequelae, several factors can affect whether sequelae are reported in relation to IMD. For example, some sequelae that receive high attention are rare in IMD survivors e.g., palsy, non-specified disruptive impulse control and conduct disorders, vegetative state, and substantial post-traumatic stress disorders (PTSD). Moreover, age substantially impacts the diagnosis of sequelae e.g., PTSD and depression may be easier to diagnose in adults than in young children. Some sequelae may be attributed to other medical conditions or may vary greatly in severity, resulting in a variable disease impact e.g., other/non-specific anxiety disorders, conduct disorder, autism, depression and oppositional defiant disorder. A number of sequelae are not typically assessed in studies, including pain, dysesthesia and aesthetic sequelae related to skin scarring.

In view of the comprehensive map, the impact of IMD on HRQOL, on other household members, survivor dependency and potentially important financial consequences where productivity of patients and caregivers is impacted [14] was usually not adequately captured in studies assessing psychological sequelae. Thus, underreporting of psychological/behavioural sequelae concerns both patients and family members. As such there is a discrepancy in the depiction of these sequelae in the literature versus clinical experience. The questions raised highlight the scarcity of available evidence in IMD survivors and the need to conduct further research to better grasp the lifelong impact of sequelae on patients and family members.

A wide range of sequela proportions were reported across studies in some cases (Table S2.2), even among high-rated studies (defined by high SIGN quality rating and high number of reported sequelae) (Table S2.3). Better strategies are, therefore, needed to select and justify the chosen point estimates and reduce uncertainty resulting from highly variable sequelae incidence. The strategy selected in this analysis prioritised high-quality studies with the largest number of sequelae outcomes reported. Robust estimates of sequela proportions may be best achieved by increasing the sample sizes, although this remains a major challenge for uncommon diseases such as IMD. In this analysis, it was not possible to pool data due to heterogeneity across study populations and designs. Future analyses may give more weight to findings from controlled studies, such as the case-control study by Viner et al. (2012) [19] and Guedes et al. (2021) [12]. Many studies lacked healthy control groups, which made interpretation of subtle differences in neurological sequelae difficult e.g., an intelligence quotient (IQ) loss below 10 points is likely to evade analysis in many studies.

Finally, regarding the EE-relevant list of sequelae, more specific definitions should be considered for neurological and neuropsychological disorders including depression. For example, the term depression covers a broad range of disease severity with variable, and often ill-defined, economic impact, in particular if it occurs with other sequelae. The sequelae considered for EE represent an important burden for IMD survivors, with clinical observations suggesting a potentially significant proportion of patients suffer from these conditions likely to lead to higher medical costs and require long-term care. The clinical evaluation also identified EE-relevant sequalae that were not included in previous EE, but which may have high or long-term costs or an important HRQOL impact, e.g., pulmonary conditions and learning disabilities.

In clinical practice, multiple organs are found to be affected both in the acute phase of IMD or over the lifetime of some survivors. The numbers of IMD patients with multiple sequelae seems, however, to be inadequately reported or quantified in OS. In a study by Darton et al. (2009) [35], nearly 30% of patients with IMD were reported to have physical and/or neurological sequelae, based on a limited list of sequelae considered i.e., defined as the need for dialysis, limb loss, skin graft, need for audiometry or visual impairment. In a US managed care population, Karve et al. (2011) [36] identified 34% of IMD patients with sequelae based on a list of up to 15 possible sequelae investigated, whereas Davis et al. (2011) [37] identified 41% of IMD patients with sequelae investigating up to 18 possible physical/neurological sequelae [4]. These studies did not consider the full range of physical/neurological and psychological/behavioural sequelae of IMD. Based on our analysis, the sum of all sequelae proportions is over 40%, but given the scarce evidence and underreporting, it is difficult to speculate if this number represents the proportion of survivors with at least one sequela, however, this does suggest a potentially high prevalence of sequelae in survivors. This is supported by the recent UK case-control study which found that 43.8% of IMD cases developed at least one sequela, compared with 23.1% of controls [12]. Further investigations should focus on the prevalence of multiple sequelae and how to address their combined impact in EE. The sequelae proportions also need further investigation due to the limited evidence available and under-reporting of psychological sequelae.

This study has a number of limitations. Comparing results across studies was challenging due to variations in study design, disease manifestation, age of subjects, study population investigated, time points and length of follow-up, as well as definitions and types of assessment used for sequelae. This heterogeneity resulted in a wide range of reported sequelae proportions and made pooling of data difficult. A more sophisticated approach is needed in future for separate analysis of data in children and adults e.g., where countries may consider different age groups for vaccination. As observed in the UK, vaccination of children and adolescents has reduced IMD incidence in these age groups, but adults are still affected, with significantly higher risks versus controls of mortality, driven by cases over 25 years old, and of severe sequelae, by cases over 50 years old [12]. The sequelae reported in each study varied considerably. Thus, additional criteria based on SIGN study quality rating and number of outcomes per study were considered, however there is still room for bias in the outcomes reported. Further statistical analysis should consider either applying standardisation methods (matching propensity analysis) to correct for patient characteristics in the study populations, or bootstrapping approaches to confirm sequelae trends where data is limited (although this alone may impose many challenges) or should focus on case-control studies. Additionally, to better understand the natural history of the disease, studies may consider analysing and applying time to follow up data to track sequelae development or confirm sequelae distribution over patients’ lifetime, or conduct sub-group analysis to determine potential risk factors (e.g., age [12]). Some generic definitions were used to group non-specific sequelae in this analysis, as not all studies clearly defined specific sequelae.

The comprehensive list of sequelae, including assumptions of which conditions are included in each definition, may be a helpful classification to harmonise reporting in future studies and allow for comparisons across studies. This study has highlighted the importance of not underestimating the burden of psychological sequelae on patients and caregivers. During the COVID-19 pandemic, caregivers experienced significantly more mental health issues than non-caregivers, highlighting some of the difficulties families face in managing chronic conditions [38]. This should be accounted for in future studies, especially when considering multifactorial disease burden i.e., COVID-19 in IMD survivors could increase the spillover burden on caregivers. While data on co-infection of IMD and COVID-19 are rare [39], a recent study in the UK in patients with invasive pneumococcal disease (IPD) co-infected with COVID-19 found the risk of death increased significantly in co-infected patients compared to IPD or COVID-19 alone [40]. During the pandemic, there was a significant decline in invasive diseases transmitted by respiratory droplets, including IMD, likely as a result of containment measures put in place to prevent SARS-CoV-2 transmission. Several of these deadly invasive diseases are vaccine-preventable, and continued vaccination can help prevent a potential increase in burden once containment measures are lifted [41, 42]. The restrictions also resulted in disruptions to vaccination provision, including childhood vaccination for IMD in some countries [41]. UK surveillance data has shown an increase in serogroup B IMD cases in the months following easing of social distancing measures, and primarily among adolescents, who are not vaccinated with 4CMenB [43].

This study comprehensively reviewed existing literature with a clinical evaluation of IMD sequelae. The results demonstrate the considerable burden that IMD survivors may experience due to physical/neurological and psychological/behavioural sequelae, and the need for additional research to quantify the impact. The full sequelae impact has been previously underestimated in EE, due to challenges in synthesising scarce evidence, and underreporting of psychological/behavioural sequelae. Comprehensive consideration of these factors (e.g., underreported sequelae, multiple sequelae, burden beyond the patient) is warranted when assessing IMD burden and potential interventions to reduce it, such as vaccination programs.