Studybox Research Guide
Planning IVD Clinical Studies Around Respiratory Season
A prospective flu, RSV or COVID-19 test study has to be enrolling before the virus arrives, so the season sets the schedule, not the sponsor.
Field notes
Guide
A flu test clinical study timeline is set by the virus, not the sponsor. Influenza and RSV in the United States usually circulate in fall and winter, and CDC reports that flu activity has most often peaked between December and February. A prospective study that needs fresh positive specimens has to be enrolling across that window, so planning works backwards from the expected season onset, with sites enrolling before regional activity starts to rise.
When do U.S. flu, RSV and COVID-19 seasons start, peak and end?
Flu and RSV follow a fall-to-spring pattern with a winter peak. COVID-19 has peaked in both summer and winter. The exact weeks shift every year and by region, so treat the patterns below as planning ranges, not dates.
Influenza. CDC’s flu season page states that “flu season usually occurs in the fall and winter” and that “most of the time flu activity peaks between December and February.” Over the 1982–83 through 2023–24 seasons, CDC’s tally of peak months shows February most often (17 seasons), followed by December (7), January (6) and March (6). A February peak is the most common outcome, not a safe assumption. CDC’s 2026–2027 season outlook says influenza hospitalizations “have typically peaked between December and February nationally, but timing within this range is less predictable.”
RSV. CDC’s RSV surveillance page says “RSV season typically begins in the fall, peaks in the winter, and ends in the spring in most of the United States.” Using pre-pandemic NREVSS data (2017–2020) for the HHS regions excluding Alaska, Florida and Hawaii, CDC defines onset as the first of two consecutive weeks with RSV PCR positivity at or above 3%. On that definition, onset ranged from mid-September to November, the peak from November to February, and the offset from early February to May. CDC notes that Florida “has an earlier RSV season onset and longer duration than most regions of the country,” and that RSV timing was disrupted by the COVID-19 pandemic before recent NREVSS data suggested a return to pre-pandemic patterns. The 2026–2027 outlook adds that the national RSV peak has typically fallen in late December or early January, although it came later in 2025–2026.
SARS-CoV-2. COVID-19 does not follow a single winter season. CDC’s 2026–2027 outlook states that hospitalizations “have varied seasonally, with peaks occurring in both summer and winter months,” and that fall and winter hospitalizations “typically have peaked in late December or early January nationally.” For a study, that means two possible enrollment windows a year, each less predictable than flu or RSV.
| Virus | Typical U.S. pattern (CDC) | Planning implication |
|---|---|---|
| Influenza | Fall–winter; peak most often December–February, February most common | Be enrolling by late fall; the peak week is not predictable in advance |
| RSV | Onset mid-Sept to Nov, peak Nov to Feb, offset early Feb to May (pre-pandemic, excluding AK, FL, HI) | Earlier onset and longer season in Florida |
| SARS-CoV-2 | Summer and winter peaks; winter peak typically late Dec–early Jan | Two windows a year; summer waves can be used |
Which CDC data sources should you watch?
Three CDC systems are useful, and they measure different things:
- FluView is CDC’s weekly U.S. influenza surveillance report, with regional data through FluView Interactive.
- NREVSS is a laboratory-based system. Participating labs report weekly test counts and positives, and the dashboard shows NAAT/PCR positivity nationally and by HHS region for SARS-CoV-2, RSV, influenza and other respiratory viruses. Percent positivity by region is the most direct signal of whether your sites are about to see positives.
- RESP-NET tracks laboratory-confirmed hospitalizations for influenza (FluSurv-NET), COVID-19 (COVID-NET) and RSV (RSV-NET) in selected areas. Hospitalizations reflect severe illness rather than outpatient test positivity, so they are more useful for reviewing past seasons than for deciding when to open an outpatient site.
Past seasons of regional NREVSS positivity are the best public guide to when positives are likely to appear in a given part of the country.
Why is prospective positive enrollment tied to the season?
Because FDA recommends fresh, prospectively collected specimens from symptomatic patients for clinical performance, and those patients only show up in large numbers while the virus is circulating.
FDA’s guidance Establishing the Performance Characteristics of In Vitro Diagnostic Devices for the Detection or Detection and Differentiation of Influenza Viruses (July 2011) recommends prospective clinical studies for every claimed type, subtype and specimen type. It asks sponsors to test “fresh specimens collected from patients suspected of having an influenza infection who have been sequentially enrolled in the study (all-comers study),” and recommends collecting the sample “within three days of the onset of influenza-like symptoms.” It also states that frozen archived specimens “are not recommended for studies to calculate clinical sensitivity or specificity.”
For influenza antigen tests, the requirements are now codified. Under 21 CFR 866.3328(b)(1), when the comparator is an FDA-cleared nucleic acid test or another accepted non-culture method, positive percent agreement for influenza A and B must have a point estimate of at least 80% with a 95% confidence interval lower bound of at least 70%, and negative percent agreement must be at least 95% with a lower bound of at least 90%. These codified special controls (2017) replace the lower antigen sensitivity targets in the 2011 guidance. For nucleic acid-based tests, the 2011 guidance recommends at least 90% sensitivity with a lower bound above 80%.
The lower bound depends on the number of positives, which only accrue while the virus is circulating. The table below shows Wilson score 95% lower bounds (z = 1.96, calculated with Python) for an antigen test at two observed agreement levels:
| Comparator positives | Device positives | Observed PPA | Wilson 95% lower bound | Meets ≥ 70% lower bound? |
|---|---|---|---|---|
| 40 | 34 | 85.0% | 70.9% | Yes, barely |
| 60 | 51 | 85.0% | 73.9% | Yes |
| 80 | 68 | 85.0% | 75.6% | Yes |
| 60 | 48 | 80.0% | 68.2% | No |
| 80 | 64 | 80.0% | 69.95% | No |
| 100 | 80 | 80.0% | 71.1% | Yes |
The interval method in your statistical analysis plan should be agreed with FDA, and an exact (Clopper-Pearson) interval gives slightly lower bounds. Either way, a device that performs near the 80% threshold needs many more positives than one at 85%.
Positives per type also scale enrollment. If comparator positivity among enrolled symptomatic subjects is 10%, then 60 influenza A positives means about 600 enrolled subjects. At 5% it means about 1,200. These rates are illustrative. Real positivity varies by week, site, age mix and which subtype dominates. When one type or subtype dominates a season, the others can be scarce. The 2011 influenza guidance anticipates this case (“when the prevalence of influenza of a particular type or subtype is unusually low”), as discussed below.
How far before the season should you start planning?
Work backwards from the earliest expected regional onset in your target regions. Everything that has to happen before the first subject is enrolled sits upstream of a date the virus chooses.
The dependencies are the same in every program: agreeing the study design with FDA (a Pre-Submission is the usual route), final protocol and statistical analysis plan, IRB approval, site contracts, device and comparator supply, and site activation and training. Each one can slip, and every slip comes out of the enrollment window, not out of the season. A study that misses the season entirely usually slips by about a year, not a few weeks. Our guide to IVD study startup delays covers the usual causes.
For a dual 510(k) and CLIA waiver program, the CLIA waiver guidance adds its own enrollment design. It recommends prospectively collected specimens and says “specimens should be collected from consecutive patients over one month,” while allowing that “depending on the specific clinical site, the prevalence of the disease, or other factors, it may be appropriate to limit consecutive enrollment to two (2) weeks.” In a respiratory study, those consecutive-enrollment periods have to fit inside the weeks when positives are actually showing up. See dual submission studies for how the two study arms fit together.
How many regions do you need to catch the peaks?
Enough to hedge against a late or uneven season. FDA’s influenza guidance recommends a minimum of three geographically diverse sites, and in practice most programs use more, spread across regions whose seasons tend to start at different times.
The 2011 influenza guidance recommends that sponsors “collect specimens and conduct your studies at a minimum of three geographically diverse facilities, one of which may be in-house,” and that point-of-care studies include sites “representative of non-laboratory settings where the device is intended to be used.” Three is a floor for representativeness, not an enrollment plan.
Season timing is uneven across the country. CDC’s RSV data show onset ranging from mid-September to November across HHS regions, with Florida earlier and longer. Influenza can rise in one region weeks before another. A site portfolio that covers several regions does three things:
- Earlier positives. Early-onset regions produce positives while others are quiet.
- Less exposure to a weak local season. Other regions can cover a mild one.
- Better population mix. The influenza guidance asks for positives from each age group, with results stratified by age, so pediatric and adult settings both matter.
Site readiness matters more in respiratory work than almost anywhere else, because a site that activates after the peak contributes very few positives. Our guide to what “pre-qualified” should mean for a site covers how to check it.
What can you do when positives run short?
Options include extending into a second season, supplementing with prospectively collected archived specimens, and, for some purposes, contrived specimens. Each has limits in FDA guidance, so agree the plan with FDA before you need it.
| Strategy | What FDA guidance says | Limits and considerations |
|---|---|---|
| Continue into another season | Fresh, prospectively and sequentially enrolled specimens are the recommended basis for clinical performance (2011 influenza guidance) | Most defensible option; adds about a year; strain mix may change between seasons, so plan whether data will be pooled |
| Prospectively collected archived specimens | Allowed “during an influenza season when the prevalence of influenza of a particular type or subtype is unusually low … to supplement the fresh prospectively collected specimens” (2011 influenza guidance) | Must be collected sequentially from all eligible patients between two pre-set dates, not pre-selected by known result; stored frozen (e.g., −70 °C); fresh-frozen equivalency study needed |
| Archived or surrogate samples in a CLIA waiver study | ”It may be appropriate to supplement prospective patient samples with archived samples”; if those are unavailable, surrogate samples “such as individual spiked or diluted patient samples” (2020 CLIA waiver guidance) | “In general, archived or surrogate samples should not comprise greater than one third of the total study samples”; must be individual samples, not pool aliquots; matrix must match; use must be justified |
| Contrived (spiked) specimens | Accepted for the fresh-versus-frozen equivalency study: “Either clinical or contrived specimens may be used,” prepared by spiking cultured virus into negative clinical matrix (2011 influenza guidance) | Not a substitute for clinical sensitivity on their own; for rare or emerging strains FDA notes evidence may come from analytical rather than clinical studies; contact FDA |
| Enriched or retrospective collections | The 2007 Statistical Guidance notes enriched studies may make results “inappropriate for pooling,” and that claims from archived collections “may or may not be appropriate” | Watch for spectrum bias; consult FDA before pooling |
How prospectively collected archived specimens work. The influenza guidance defines them as specimens “collected sequentially from all patients meeting study inclusion criteria and representing the assay’s intended use population (i.e., not pre-selected specimens with known results) … between two pre-determined dates (e.g., from the beginning to the end of one flu season), so there is no bias and the prevalence is preserved.” If the archived specimens were cultured before freezing, test them only with the investigational device and compare against the original culture results. If not, thaw them and test with both an acceptable NAAT comparator and the device. The guidance also recommends a fresh-frozen equivalency study showing positive agreement of at least 95% with a 95% lower bound above 90%, using at least 60 swab samples and 60 wash or aspirate samples.
This backup has to be planned ahead: sequential collection and frozen storage under an IRB-approved protocol, set up during the season itself.
What does the annual strain-reactivity requirement mean for influenza tests?
For influenza antigen tests, 21 CFR 866.3328(b)(3) requires annual analytical reactivity testing against contemporary strains. Results go into the labeling by July 31 each year. This is a recurring post-clearance obligation, separate from the clinical study.
The special control reads: “Annual analytical reactivity testing of the device must be performed with contemporary influenza strains.” Under § 866.3328(b)(3):
- FDA, in consultation with CDC, identifies the strains to test. They are sourced from CDC or an FDA-designated source.
- Testing must follow “a standardized protocol considered and determined by FDA to be acceptable and appropriate.”
- “By July 31 of each calendar year, the results of the last 3 years of annual analytical reactivity testing must be included as part of the device’s labeling” (or every year since clearance if the device is newer). Results go in a table, either in the package insert or on a public web page linked prominently from the labeling and the manufacturer’s site.
Section 866.3328(b)(4) adds a separate clock for declared influenza emergencies: testing within 30 days of FDA’s notice that characterized samples are available, and labeling of results starting within 60 days.
The codified annual-reactivity text is specific to influenza antigen tests under § 866.3328. Molecular influenza assays are classified differently. For example, § 866.3980 (respiratory viral panel multiplex nucleic acid assays) relies on special controls guidance documents, and § 866.3332 covers reagents for specific novel influenza A viruses. Check the regulation and product code for your device, and any conditions in your clearance, rather than assuming the antigen-test schedule applies. For SARS-CoV-2 devices, § 866.3981 and § 866.3982 have their own special controls. Both call for clinical studies with prospective samples; § 866.3982 specifies “a prospective multisite clinical study.”
For planning, an influenza antigen test carries a fixed yearly deliverable after clearance; budget for it before launch.
What do respiratory studies look like in practice?
Several respiratory programs Studybox has supported follow this pattern. A dual 510(k) and CLIA waiver study for a point-of-care molecular Flu A/B test enrolled prospectively at 16 intended-use sites within a single influenza season, using fresh nasal swabs from children and adults. A dual study for a point-of-care molecular RSV test ran as a prospective multi-site U.S. study during one RSV season, with nasal swabs tested within one hour of collection. A point-of-care antigen test for SARS-CoV-2 and influenza A/B shows why multiplex targets are harder: each analyte has its own positive target and its own circulation pattern. A dual study for a point-of-care molecular Strep A test applied the same approach to a bacterial pharyngitis target: a nine-site prospective study on fresh throat swabs at point-of-care sites.
How Studybox approaches this
Studybox Research is an IVD-only CRO and clinical site network with more than 100 pre-qualified U.S. sites in urgent care, physician office and point-of-care settings. For respiratory programs, a larger pool of ready sites gives more options for placing enrollment where regional positivity is rising. Typical study activation is about four weeks, compared with a 3–6 month industry average, which leaves more room to finalize protocols closer to the season.
The team has supported more than 55 FDA regulatory clearances across 510(k), dual 510(k)/CLIA waiver, OTC and EUA pathways. Prospective specimen collection under a pre-approved IRB protocol can launch in as little as one week. That is useful for building a prospectively collected specimen archive during the season. To plan a respiratory study around the coming season, see our CLIA waiver studies and 510(k) studies pages or contact us.
Studybox Research Guide October 5, 2026