The National Vaccine Injury Compensation Program (VICP), authorized by the National Childhood Vaccine Injury Act of 1986, as amended, and administered by the Health Resources and Services Administration (HRSA), has been in operation since 1988. It is a no-fault alternative to the traditional tort system to adjudicate petitioner (i.e., individuals filing claims) claims of alleged vaccine-related injury [1]. The program covers all vaccines routinely recommended for children and/or pregnant women that have an excise tax placed on them and are added to the Vaccine Injury Table by the Secretary of the Department of Health and Human Services. Petitions may be filed with the VICP regardless of the age of the recipient; for example, because influenza vaccine is routinely recommended for children and is subject to an excise tax, it is a vaccine covered under the VICP and both children and adults who receive the influenza vaccine are eligible to file claims alleging a related injury or death.

The Vaccine Adverse Event Reporting System (VAERS), also authorized by the National Childhood Vaccine Injury Act of 1986, is a national spontaneous reporting system for adverse events following vaccines. VAERS, which is co-administered by the CDC and FDA, is used to conduct post-licensure vaccine safety monitoring and is separate and unrelated to the VICP [2]. A report to VAERS does not constitute or trigger a VICP petition, nor does a petition to VICP necessarily lead to a report to VAERS.

In 2010, physicians from the VICP published a case series studying 13 petitioners with severe arm and shoulder pain and reduced range of motion following vaccination; signs and symptoms lasted greater than six months and radiographic and surgical findings were consistent with injury to musculoskeletal structures of the shoulder [3]. The authors introduced the term shoulder injury related to vaccine administration (SIRVA) to describe these cases. SIRVA is thought to occur when vaccine is injected above or through the deltoid into underlying non-muscular tissues, such as the bursae, ligaments, and tendons [3]. Specific impairments include bursitis [4–8], bone erosion [8], and damage to the rotator cuff and bicipital tendons [4,6]. The authors, including a rheumatologist, hypothesized that an inflammatory response to vaccine contents in tissues other than the muscle may be causative [4,5,8]. Based on case series descriptions of SIRVA and SIRVA-like outcomes, a 2012 Institute of Medicine report concluded that deltoid bursitis can be causally associated with vaccination [9].

Through the federal rule-making process, VICP proposed (in 2015) and officially added (in 2017) SIRVA to the Vaccine Injury Table (Table) [10,11]. To gain entitlement to compensation under the VICP, petitioners must establish a vaccine-related injury or death has occurred, either by proving that a vaccine actually caused or significantly aggravated an injury (‘‘causation-in-fact”) or by demonstrating the occurrence of what is referred to as a ‘‘Table injury.” That is, petitioners may show that the vaccine recipient suffered an injury of the type enumerated in the Table corresponding to the vaccination in question and that the onset of such injury took place within the time period also specified in the Table. If so, the injury is presumed to have been caused by the vaccination, and the petitioner is entitled to compensation (assuming that other requirements are satisfied), unless it is found that the injury was caused by some factor other than the vaccination. Prior to SIRVA’s addition to the Vaccine Injury Table on March 21, 2017, VICP reviewed, adjudicated, and, where appropriate, conceded petitioner claims of SIRVA-like injuries based on causation-in-fact. During the time period from the 2010 case series publication [3] to the addition of SIRVA to the Vaccine Injury Table, alleged SIRVA petitions to VICP had been increasing substantially, especially for influenza vaccine [12].

In 2017, HRSA and the Centers for Disease Control and Prevention (CDC) initiated a joint scientific review to evaluate clinical characteristics of SIRVA petitions to the VICP since the time this condition was described by Atanasoff and colleagues [3].

Our review of SIRVA petitions used information from the Injury Compensation System, a VICP administrative database that is maintained by HRSA as a part of routine program administration. The process for identifying, reviewing, categorizing, and analyzing SIRVA petitions is detailed below.

Petitions to the VICP for SIRVA and SIRVA-like injuries have increased every year since the initial case series of petitioner claims for SIRVA to VICP was published (Fig. 1), rising from 2.5% of total petitioner claims in 2011 to 41.9% in 2016 (Fig. 2). Our initial query (using the broadest criteria specified in the Methods) yielded 913 medical reports during the analytic period, 105 of which VICP clinician reviewers determined to be conditions other than SIRVA. Of the remaining 808, the Medical Analysis Branch recommended 476 petitions for concession for SIRVA as causation-in-fact; we included these cases in the final review (Fig. 3). Demographics and characteristics of cases are summarized in Table 1. Median age of the cases was 51 years (range 16–92), 82.8% were women, and median body mass index (BMI) was 25.1 (range 17.0–48.9; normal BMI is 18.5–24.9).

Petitioner claims to the VICP for alleged SIRVA and SIRVA-like injuries have risen dramatically since the initial case series of petitioner claims for SIRVA to VICP was published in 2010 [3]. In the time period between that publication and the addition of SIRVA to the Vaccine Injury Table, the VICP recommended concession as causation-in-fact for 58.9% of the 808 claims for SIRVA or related shoulder injuries.

We found many similarities between our case series and other case reports described in the literature, including a predominance of women and adult patients [3,6]. The median age of our cases was 51 years, with only one, a teenager, below age 18 years. Children and adolescents receive many vaccines, and children over the age of three years are recommended to be vaccinated in the deltoid region for intramuscular injections. In a study of bone strikes during vaccination of children, Lippert and Wall determined that use of the recommended needle lengths by age can lead to over-penetration of the deltoid muscle in 10.6–61.3% of cases [13]. Despite this potential risk, children are rarely the subject of petitioner claims for SIRVA in the VICP. In general, soft tissue shoulder injuries, such as ligament disruptures and bursitis, are uncommon in pediatric populations [14].

Initial research on potential vaccination-related shoulder injuries suggested that people with lower BMI might be at greater risk of over penetration of the deltoid muscle and therefore at increased risk for injury [4,15]. However, later case series analyses indicated that BMIs generally ranged from normal to obese [3,16]. We did not find evidence of low BMI preponderance in our cases; the median BMI was in a range that CDC defines as ‘‘overweight” [17] and the distribution of BMI of our cases is similar to the BMI distribution of the United States general population.

Most (84.0%) of our cases involved inactivated influenza vaccine (Table 2). This is the most common vaccine administered in the United States, with approximately 150 million doses distributed annually [18] and the majority of doses administered in adults. As such, it is not surprising that inactivated influenza vaccine was the most frequently implicated vaccine. In addition, the most common place of vaccination was in pharmacies or stores (Table 2), which is consistent with SIRVA reports received by the Vaccine Adverse Event Reporting System (VAERS) of atypical shoulder pain and dysfunction following inactivated influenza vaccine [19].

SIRVA can be broadly defined as pain that begins within 48 h of vaccination and reduced range of motion [11]. This definition is sensitive enough to include almost all musculoskeletal injuries that can result from injection, but such a broad definition makes it difficult to describe SIRVA as a discrete clinical entity. Consistent with this lack of specificity, we observed that initial diagnoses at the first clinical exam varied (Table 4), with shoulder pain being the most common finding and most common initial diagnosis, followed by rotator cuff problems and bursitis. Common findings on MRI included tendonitis/osis/inopathy, rotator cuff tears and bursitis (Table 5). Similar findings have been previously described in case reports of shoulder injury following vaccination [16,20]. Although less common, bone edema and necrosis was also evident among our cases, which has also been reported in the literature [8,21]. A general clinical or radiographic finding among our cases was evidence of injury to internal structures, primarily, but not limited to, the tendons of the rotator cuff and bursae. MRI findings in our cases should be interpreted with caution and not be equated with prima facie evidence of a causally associated injury. Such findings are not uncommon among people in the age group of our cases, independent of vaccination (i.e., shoulder pathology is common in middle age and old age) [22]. In addition, corticosteroid injection into and around the joint may also be responsible for some MRI findings [23].

Less than a quarter (24.3%) of cases reported that their symptoms had resolved or largely resolved by the last clinic appointment available in the medical records. However, all petitioners in the analysis were seeking financial compensation for an alleged SIRVA injury and had a disincentive to acknowledge resolution of symptoms (a possible perceived barrier to receiving compensation), so this finding may not be unexpected. Medical interventions were common; the majority of cases were referred for physical or occupational therapy or received corticosteroid injections into the affected joint and around one-third had surgery. Due to potential bias in self-reporting of recovery status, it is not possible to evaluate either the impact or relative impact of specific treatments for SIRVA.

With 476 petitioner claims, this is the largest case series of suspected SIRVA and SIRVA-like injuries to date. The VICP clinician reviewer assessment documented in medical reports allowed us to exclude cases that petitioned for compensation, but whose injuries were unlikely to have been SIRVA, with reviewer diagnoses including brachial neuritis, cervical radiculopathy, complex regional pain syndrome, arthritis, and others. The medical reports (summaries) that served as our data source were based on comprehensive medical records submitted to the VICP as part of the claims adjudication process. During the medical report abstraction process, our physician-nurse adjudication committee was in frequent consultation to adjudicate issues when questions and disagreements arose, which minimized inter-reviewer variation.

Our review has limitations. All petitioners submitting claims to the VICP are seeking monetary compensation for perceived injuries from vaccination, which may impact their use of the healthcare system (i.e., frequency and type of care), reported severity of symptoms, and statements on recovery status. Because the VICP only covers certain vaccines, some vaccines, such as pneumococcal polysaccharide and recombinant zoster vaccines, are not included in the VICP; thus, SIRVA injuries related to such vaccines would not be captured in our case series. In addition, SIRVA is different from other compensable injuries in the VICP. Most of the other Vaccine Injury Table injuries are rare conditions, whereas shoulder injuries are common, accounting for approximately 1% of all primary care visits in the United States [24]. As the VICP system is designed to give claimants the benefit of doubt, it is possible, if not probable, that cases of shoulder injuries that were not related to vaccination (i.e., pre-existing or coincidentally occurring) were included in petitioner claims and conceded by the VICP. Furthermore, since we have neither denominator data nor an unvaccinated comparison group, we cannot determine the risk of SIRVA or individual risk factors. Finally, our abstraction documents were medical reports from the VICP, which are legal documents and not designed for use in epidemiologic analysis.

Petitioner claims for SIRVA to the VICP increased substantially from 2010 to 2016, following the initial SIRVA case series publication from the VICP and the open vetting process for adding SIRVA to the Vaccine Injury Table. With hundreds of millions of doses of adult vaccines distributed in the United States in the 2000 s, including 100–150 million doses of influenza vaccines distributed annually in recent influenza seasons, conceded petitioner claims for SIRVA in the VICP still appear to be relatively uncommon. However, true SIRVA is likely preventable. Injection too high on the arm could be a factor due to the risk of injecting into underlying non-muscular tissues (bursa, tendons, ligaments, and bones). Healthcare providers and clinical practice coordinators should ensure that all staff members who administer vaccines receive proper training on vaccination technique. When injecting into the deltoid, care should be taken to ensure that the injection is placed in the thick, centrally located portion of the deltoid muscle, away from the upper third of the deltoid where the risk of over penetration into underlying structures of the shoulder is greatest.

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention, the Health Resources and Services Administration, or the U.S. Department of Health and Human Services.