Common patterns

Overview #

Across primary-series, booster, bivalent, and pediatric MMWR briefs, everyday reactogenicity rates often track the v-safe dataset—while the serious-event story does not. The same few habits repeat: hospital checkboxes without the standard hospital MedDRA term; medical care narrated as “unrelated”; call-center follow-up that frequently fails; cardiac and seizure coding left off the v-safe Results page next to “no myocarditis” or “serious adverse events are rare”; and absolute hospital-empty claims that overread a week-1 table cell.

Each pattern lists example papers (DOI, title, date, authors). Author names link to the affiliation legend at the bottom (SearchBuilder-enabled; names normalized across papers). Pattern write-ups include at least one concrete registrant you can open in the v-safe viewer. For hospital and ER under-coding, see the long-form pages Hospitalization without Hospitalisation PT and ER checkbox without emergency-care PT.

1. Hospitalization checkbox without Hospitalisation MedDRA #

Full deep dive → (why missing PT_NAME matters, verbatim quotes, dozens of registrant cases, searchable grid).

On the week-1 health check-in, a parent or registrant can tick Hospitalization (often alongside ER). That checkbox is a direct claim that the vaccinated person was admitted. Coders are also supposed to attach MedDRA preferred terms. The standard term for that setting is Hospitalisation. In reconstructed cohorts, that PT is often missing on the same response that carries the hospital checkbox—sometimes only about one in five hospital-checkbox rows get it—while symptom codes (appendicitis, pneumonia, chest pain, anaphylaxis) still appear. Free text may describe the admission clearly; the coded inpatient flag still disappears from tallies that rely on the PT.

Example. In the 5–11 Pfizer booster brief (mm7133a3), registrant DQ6-14633-25689 checked ER + Hospitalization the day after dose 3 and was coded Chest pain; Dyspnoea; Heart rate increased—with no Hospitalisation preferred term. Call-center follow-up closed as Unsuccessful. That same paper’s Results claim “No children received care at a hospital after third dose vaccination” (see also pattern 5).

Nearly every HTML report documents this; later-week rows in pediatric papers often lack the PT even when week-1 coding occasionally succeeds.

2. Soft “unrelated / declined / error” disposition #

When a brief admits hospitalizations or medical care, the sentence readers remember is usually that parents said care was unrelated to vaccination, declined a VAERS report, or checked hospitalization in error. Those counts are taken only among people with call-center (or free-text) follow-up information—often a minority of care-seekers—and the paper rarely states how many follow-ups closed as Unsuccessful. The soft disposition becomes the entire story of serious care.

Example. The ≥12 bivalent booster brief (mm7144a3) narrates 55 hospitalizations, then “29 unrelated / 13 VAERS / 3 declined” among those with information—while Table 2 prints zeros (pattern 5). Among hospital reporters in reconstruction, call-center outcomes are heavily mixed (roughly Successful ≈ Unsuccessful). One hospital-checkbox registrant with no Hospitalisation PT and Unsuccessful follow-up is RKX-30112-69199 (open in the viewer; see the paper’s hospital grid). The 5–11 bivalent brief (mm7202a5) similarly leads with “35 of 37 unrelated” while roughly 20 of 54 week-1 medical-care reporters in that extract never closed successfully.

Classic form: “29 unrelated / 13 VAERS / 3 declined” (bivalent adults) or “35 of 37 unrelated” (5–11 bivalent).

3. Failed call-center follow-up (GDIT Unsuccessful) #

Methods sections say CDC’s v-safe call center reaches out when someone reports medical care or hospitalization. In reconstructed cohorts, GDIT outcomes marked Unsuccessful are often roughly half—or more—of hospital or medical-care reporters. Papers still write as if the subset with “additional information” (unrelated / VAERS / error) represents how care-seeking was resolved. Sometimes the clearest hospital case in the extract itself closes as Unsuccessful.

Example. In the 6 months–5 years primary-series brief (mm7135a3 / hospital section), registrant BBU-44168-22141 has a week-1 Hospitalization checkbox with Respiratory distress coding and no Hospitalisation PT—follow-up outcome Unsuccessful. In the 12–17 booster brief (mm7109e2), 8W7-32430-24116 is a hospital-path row whose free text notes discharge, also closed Unsuccessful—while Results spotlight a single “new onset migraine” hospitalization anecdote.

4. Chest pain / dyspnea unpublished next to “no myocarditis” #

Abstracts and Discussions reassure with no myocarditis (or VAERS-only myocarditis counts) and lines like “serious adverse events are rare.” The v-safe Results page almost never publishes post-dose Chest pain, Dyspnoea, Palpitations, or related preferred-term counts—even when dozens to hundreds of registrants in the reconstructed cohort carry those codes, and free text names heart pain, ER work-ups, or angina. Readers are steered to VAERS cardiac review while the active surveillance stream’s cardiac symptom coding stays offstage.

Example. After the 5–11 primary series (mm705152a1), registrant 3EC-31970-76337 has MedDRA Chest pain and free text “Headaches, chest pains and increasing stomach pain.” RV7-18261-94082 is coded Chest pain; Palpitations with free text naming heart palpitations. Neither appears in the paper’s v-safe Results narrative next to its soft hospital / rare-SAE framing. In the 5–11 homologous booster brief (mm7133a3), EVF-41977-66694 wrote “Heart palpitations and chest pain beginning July 2, 2022” beside a VAERS “no myocarditis” claim.

5. Absolute “no hospital” claim or Table zeros #

Some briefs print Hospitalization = 0 in every Table 2 age column—even when the Results paragraph admits dozens of hospital reports—or write absolute prose (“no children received care at a hospital”) that readers treat as the whole surveillance story. Absolute zeros screenshot cleanly; the contradiction (narrative 55 vs table zeros, or a week-1 hospital checkbox vs “none”) does not.

Example (absolute claim contradicted). Hause et al. (mm7133a3) state: “No children received care at a hospital after third dose vaccination.” Registrant DQ6-14633-25689 checked ER + Hospitalization on day 1 after dose 3, with Chest pain / Dyspnoea / tachycardia coding and Unsuccessful follow-up (same child as in pattern 1).

Example (table zeros vs text). The ≥12 bivalent brief (mm7144a3) prints Hospitalization 0 across Table 2 while Results admit 55 hospital reports. Open any hospital-grid row such as RKX-30112-69199 to see a week-1 hospital checkbox the table never surfaces. In mm7202a5, week-1 Hospitalization = 0 matches Table 2, but Results prose “no children received hospital care” overreads that week-1 cell (see pattern 8).

mm7133a3: “No children received care at a hospital” contradicted in-cohort. mm7144a3: Table 2 all zeros vs text 55. mm7202a5: week-1 zero overread as absolute.

6. Seizures unpublished in v-safe Results #

Pediatric and toddler briefs often let VAERS name seizures among serious reports (sometimes then ruled “unrelated after review”)—while matching v-safe Seizure / Febrile convulsion preferred terms, and free text such as a first-time seizure hours or days after a dose, never appear on the active-surveillance Results page next to “SAEs rare” / “reassuring.”

Example. After a 5–11 bivalent booster (mm7202a5), registrant 3TZ-12549-33451 has free text “Cold with a fever and she had a seizure,” MedDRA including Seizure; Emergency care, and a later Hospitalization checkbox (day ~91) still without a Hospitalisation PT on that response. In the 6 months–5 years series (mm7135a3), CL6-20696-00749 wrote “Seizure. High Fever 104.7 F” with Seizure coding, and 2US-32065-49888 “Had febrile seizure on 9/5” (Febrile convulsion)—none of which surface in the v-safe Results reassurance block.

7. Deaths sidelined to VAERS-only narrative #

Early adult and adolescent briefs discuss deaths almost exclusively through VAERS review language (“under review,” “no causal association suggested,” long case lists). Meanwhile v-safe free text and MedDRA that carry Death (or cardiac-arrest pathways ending in death) rarely—if ever— appear in the paper’s v-safe Results story. Some pediatric windows truly lack a Death PT; the recurring issue is asymmetry: VAERS may name death at length while active-surveillance Death coding in the same study window stays unpublished.

Example. Adult booster monitoring (mm7107e1) includes registrant 87R-15615-42698 with MedDRA Death and free text that the person “died on 10/16/22 from a massive coronary…” (registrant asked to remove the record). That sits in the v-safe stream while death discussion in these briefs stays VAERS-centered. In the Janssen adult brief (mm7018e2), 5DU-20396-72475 has free text that the registrant passed after heart failure / blood clot, with MedDRA Cardiac failure; Death; Thrombosis. Second-booster adults (mm7130a4) include NYE-12336-57179 (pulmonary embolism / cardiac arrest pathway with Death coding)—again outside the v-safe Results pitch.

8. Later hospitalization checkboxes ignored #

Table 2 (and most v-safe reactogenicity tables) cover days 0–7 after the dose. Papers still write absolute prose—“no children received hospital care”—that readers treat as the whole active-surveillance story. After week 1, the same reconstructed cohorts often show further Hospitalization checkboxes—sometimes with Seizure, Type 1 diabetes, RSV, sepsis, or cardiac coding—and usually still without a Hospitalisation preferred term on that later response.

Example. In toddler/preschool third doses (mm7223a2), registrant C4R-14340-74428 (day ~14) has free text “He caught RSV and has been struggling with it. He's out of the hospital now…” with RSV; Hospitalisation coding. 5YE-88466-31434 later checks hospitalization with Epilepsy; Seizure; Emergency care—themes that parallel VAERS serious reports the paper dismisses after review, while these later v-safe rows stay off the Results page. In mm7202a5, the same 3TZ-12549-33451 seizure case later shows a Hospitalization checkbox (see pattern 6).

9. “v-safe has no diagnoses” while free text is coded #

Several papers tell readers that v-safe cannot identify why someone sought care or was hospitalized—or that it “does not record reason for hospitalization.” The same system stores free-text symptom and visit descriptions and MedDRA preferred terms (Myocarditis, Appendicitis, Pneumonia, and more). The Methods/Results hedge lowers expectations so unpublished diagnoses never look like an omission of data that already exists.

Example. The adolescent primary-series brief (mm7031e1) claims v-safe “does not record reason for hospitalization.” Registrant LEW-15917-28261 wrote “Hospitalized for myocarditis, and chest pain, always tired” and carries Myocarditis; Hospitalisation; Chest pain coding; WEW-50402-71064 wrote “Diagnosed with myocarditis.” In the 12–17 booster brief (mm7109e2), Results spotlight one migraine hospitalization while LH6-45059-75313 reports an appendectomy (ER + hospitalization; Appendicitis / Appendicectomy codes—still often without a clean Hospitalisation PT), never described in the v-safe Results narrative.

Author & affiliation legend #

Click an author name in any pattern table above to jump here. Names and CDC team labels are normalized across papers (e.g. with/without degrees; Response vs Emergency Response Team). Use SearchBuilder or the search box to filter by author, affiliation, or paper id.

How to read these reports #

Start from a pattern above, open a linked paper, and use the page Contents / # anchors for the section with paper quotes, plain-language explanation, and searchable evidence grids (registrant codes link to the v-safe viewer). Full analysis text also lives under vsafe-analysis/mmwr/<id>/anomalies_report.md.

← All MMWR anomaly reports