This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
SPRINT-SARI is the Short Period Incidence Study of Severe Acute Respiratory Illness. The ClinicalTrials.gov record identifies severe acute respiratory infection as the condition and lists the Australian and New Zealand Intensive Care Research Centre as the lead sponsor. The registry lists the study as recruiting, with a start date of 2016-01, a primary completion date of 2026-12, and an enrollment of 15000.
| Feature | SPRINT-SARI |
|---|---|
| Study title | Short Period Incidence Study of Severe Acute Respiratory Illness |
| Condition | Severe Acute Respiratory Infection |
| Status | Recruiting |
| Planned enrollment | 15000 |
| Study start | 2016-01 |
| Primary completion date | 2026-12 |
| Lead sponsor | Australian and New Zealand Intensive Care Research Centre |
| Sponsor type | OTHER |
| ClinicalTrials.gov | NCT02498587 |
2. Clinical Question
SPRINT-SARI is structured around a data-collection question rather than a conventional randomized treatment comparison. The registry-defined primary endpoints ask three related operational questions: how many participating sites can participate and submit data for central analysis, how complete the submitted data are, and what barriers to data completion are identified after the SPRINT-SARI study period.
Population
The registry identifies the condition as severe acute respiratory infection and lists a planned enrollment of 15000.
Intervention
The ClinicalTrials.gov record does not identify a therapeutic intervention or treatment arm in the information reported here.
Comparator
The registry information does not identify a randomized comparator or control treatment.
Primary question
Can participating sites submit centrally analyzable data, how complete are those data, and what barriers affect data completion?
3. Trial Design
The registry identifies SPRINT-SARI as the Short Period Incidence Study of Severe Acute Respiratory Illness. Its primary endpoints focus on site participation and the quality and completeness of submitted data rather than a treatment-effect comparison.
4. Endpoints
The ClinicalTrials.gov record specifies three registered primary endpoints. Each has a distinct statistical target and time frame.
| Primary endpoint | Time frame | Registry definition |
|---|---|---|
| Number of participating sites | one week | The number of sites able to participate and submit data for central analysis |
| Data Completeness | 90 days | The completeness of submitted data |
| Barriers to data submission | 90 days | Survey post SPRINT-SARI study period on barriers to data completion |
Endpoint 1: Number of participating sites
The first endpoint is defined as the number of sites able to participate and submit data for central analysis, with a time frame of one week. Statistically, this is a count rather than a conventional continuous or time-to-event outcome.
Endpoint 2: Data Completeness
The second endpoint is Data Completeness, with a time frame of 90 days. The registry defines this as the completeness of submitted data. The definition establishes the construct being evaluated, but the registry information reported here does not provide a numerical scoring formula or a prespecified threshold for classifying a submission as complete.
Endpoint 3: Barriers to data submission
The third endpoint is Barriers to data submission, also assessed at 90 days. The registry describes this as a survey conducted post SPRINT-SARI study period on barriers to data completion. This is therefore conceptually different from a treatment-effect endpoint: the statistical target is information about reasons for incomplete or difficult data submission.
5. Statistical Methodology
The three endpoints require different statistical approaches because they measure different constructs. A site count, a completeness measure, and a post-study survey should not automatically be analyzed with the same statistical method.
| Endpoint | Outcome type | Typical statistical approach |
|---|---|---|
| Number of participating sites | Count | Descriptive count of sites meeting the registry definition within the one-week time frame |
| Data Completeness | Completeness measure | Descriptive summary of completeness, with the exact analysis depending on how completeness is operationalized |
| Barriers to data submission | Post-study survey | Descriptive survey analysis; categorical responses would typically be summarized using counts and percentages |
Counts and proportions
The first endpoint is inherently a count. If the registry's operational definition is satisfied by individual sites, the primary descriptive result would naturally be the number of qualifying sites. If the total number of eligible or participating sites were also known, a proportion could provide additional context, but that denominator is not specified in the registry information reported here.
Completeness as a statistical construct
Data completeness is only statistically reproducible once the unit of completeness and scoring rule are specified. For example, completeness could be represented as the percentage of required fields populated, the percentage of expected records received, or another prespecified measure. These approaches are not interchangeable because they can produce different summaries from the same underlying data.
This formula illustrates the general statistical structure of a completeness metric. The ClinicalTrials.gov record identifies the endpoint as the completeness of submitted data but does not provide a numerical formula in the information reported here.
Survey analysis
The barriers endpoint is based on a survey conducted after the SPRINT-SARI study period. If responses are categorical, the basic analysis would typically report response counts and percentages. If multiple barriers could be selected by a respondent, each barrier would generally be summarized separately because the percentages could then sum to more than 100%.
Time frames are part of the estimand
The endpoint time frames are not merely administrative labels. A one-week endpoint measures early site participation and data submission capability, whereas a 90-day endpoint evaluates a longer period of data completeness and post-study barriers. Changing the assessment window would change the quantity being measured.
6. Planned Analysis
No formal statistical analyses are posted to ClinicalTrials.gov for the registered primary endpoints. The registry nevertheless defines the quantities to be measured, which allows their statistical roles to be described.
Participating sites
The registry will assess the number of sites able to participate and submit data for central analysis over one week. The natural primary summary is a count of qualifying sites.
Data completeness
The registry will assess completeness of submitted data over 90 days. The appropriate analysis depends on the operational definition used to quantify completeness.
Submission barriers
The registry will use a survey after the SPRINT-SARI study period to evaluate barriers to data completion. Descriptive summaries of survey responses would typically be appropriate.
Interpretive focus
These endpoints describe data-collection performance rather than estimating a randomized treatment effect.
7. Statistical Methods Explained
Why is the number of participating sites treated as a count?
The first endpoint is explicitly defined as the number of sites able to participate and submit data for central analysis. A count is therefore the direct statistical representation of the endpoint. The key issue is not selecting a sophisticated model automatically, but defining precisely which sites satisfy the endpoint criterion within the one-week time frame.
Why is data completeness different from a conventional clinical endpoint?
Data completeness describes the quality or availability of collected information rather than a biological outcome experienced by a patient. Statistical analysis therefore begins with the structure of the submitted data: what information was expected, what information was received, and how completeness is calculated. Without that operational definition, a numerical completeness result cannot be interpreted consistently.
Why would percentages be useful for the site endpoint?
A count tells the reader how many sites qualified. A percentage can additionally express that count relative to a clearly defined denominator. For example, if the number of sites assessed for eligibility were known, the proportion meeting the submission criterion could provide a standardized measure. The denominator must be defined before such a proportion is interpreted.
How would a survey of barriers normally be summarized?
Categorical survey responses are commonly summarized with counts and percentages. If respondents can identify multiple barriers, each response category can be summarized independently. The resulting percentages do not necessarily sum to 100%, because one respondent may contribute to more than one category.
Why does the 90-day time frame matter?
A completeness measure evaluated over 90 days answers a different question from an immediate submission check. A site may be able to submit an initial dataset promptly but encounter later problems with missing or incomplete information. The prespecified time frame therefore defines the observation window for the endpoint.
Why should the three endpoints not be collapsed into one statistic?
They measure different dimensions of study performance: site participation, data completeness, and barriers to submission. Combining them without a prespecified composite definition could obscure important differences between these dimensions. A site could participate successfully while still having incomplete data, for example, and the barrier survey could identify reasons that would not be visible from a site count alone.
8. Analysis Populations and Denominators
For conventional randomized trials, analysis populations such as intention-to-treat and as-treated populations are central to interpretation. The registry information for SPRINT-SARI does not identify randomized treatment groups or provide corresponding efficacy and safety analysis populations.
| Statistical question | What the registry establishes | Interpretive consequence |
|---|---|---|
| Who is randomized? | No randomized treatment allocation is identified in the reported registry information. | A randomized ITT treatment comparison cannot be defined from these data. |
| What is the treatment comparator? | No intervention or comparator treatment is identified. | Hazard ratios, risk ratios, and treatment-effect estimates are not the primary statistical framework here. |
| What is the denominator for completeness? | The endpoint is defined as completeness of submitted data. | The precise denominator depends on the operational completeness definition. |
| What is the denominator for the site count? | The endpoint is the number of sites able to participate and submit data for central analysis. | A percentage requires an additional defined denominator. |
9. Missing Data and Completeness
Missingness is particularly important in a study whose primary endpoints include data completeness and barriers to data submission. Here, incomplete information is not simply a nuisance that occurs around a conventional clinical endpoint; it is itself part of what the study is designed to examine.
Statistically, missingness can arise at multiple levels. A site may fail to submit data, a submission may contain only some expected information, or particular data elements may be absent within an otherwise submitted record. Each pattern implies a different completeness calculation and potentially a different operational response.
Site-level missingness
A site may be unable to participate or submit data for central analysis. This directly relates to the first primary endpoint.
Record-level missingness
Submitted data may contain missing information even when a site successfully submits a dataset.
Variable-level missingness
Specific required data elements may be absent within otherwise usable records.
Barrier information
The post-study survey provides a mechanism for characterizing barriers associated with incomplete data submission.
These distinctions matter because a single overall completeness percentage can conceal whether missingness is concentrated in a small number of sites or distributed across many submissions.
10. Why This Trial Matters Statistically
SPRINT-SARI illustrates a different side of clinical-trial statistics: statistical quality control for clinical data collection. The central statistical challenge is not estimating a treatment effect but defining measurable properties of participation and data quality so that the resulting dataset can support central analysis.
| Concept | How it appears in SPRINT-SARI |
|---|---|
| Endpoint definition | Three primary endpoints separately address site participation, data completeness, and barriers to data submission. |
| Count data | The number of participating sites is a discrete count. |
| Data quality measurement | Data Completeness is a primary endpoint measured over 90 days. |
| Survey methodology | Barriers to data submission are evaluated using a post-study survey. |
| Time windows | Primary endpoints use one-week and 90-day time frames. |
| Denominator definition | Interpretation of completeness and proportions depends on clearly defining expected observations or data elements. |
| Missing data | Incomplete submissions are directly relevant to the study's primary measurement objectives. |
| Central analysis | The participating-site endpoint explicitly refers to sites submitting data for central analysis. |
11. Interpreting Data Completeness
Data completeness is often reported as a simple percentage, but the percentage is only meaningful when the numerator and denominator are transparent. Consider a dataset in which each record is expected to contain several required fields. A completeness metric could count all populated required fields, or it could require an entire record to be complete before that record contributes to the numerator. Those two definitions answer different questions.
Both are legitimate statistical constructs in appropriate settings, but they are not numerically interchangeable. The precise operational definition determines what a reported completeness percentage means.
For a multicenter clinical dataset, site-level summaries can also be informative. A high overall completeness percentage can coexist with substantial variation between sites if a large volume of data comes from sites with high completeness while a smaller number of sites have substantial missingness.
12. Barriers to Data Submission
The third primary endpoint moves from measuring the extent of data-completion problems to investigating their causes through a survey conducted after the SPRINT-SARI study period.
From a statistical perspective, survey responses can be summarized descriptively before any more complex modeling is considered. The fundamental quantities are the number of respondents, the number selecting each barrier, and the corresponding percentages using an explicitly defined denominator.
| Survey quantity | Typical statistical summary |
|---|---|
| Respondents | Count of completed survey responses |
| Individual barrier | Count and percentage selecting the barrier |
| Multiple barriers | Each barrier summarized independently when multiple selections are allowed |
| Missing survey responses | Reported separately so that response denominators remain interpretable |
The statistical interpretation should distinguish between the prevalence of a reported barrier and evidence that the barrier caused incomplete data. A survey response can identify a perceived or reported barrier, but causal interpretation requires an appropriate design and analysis beyond a descriptive response frequency.
13. One-Week Versus 90-Day Assessment
The endpoint schedule creates a useful distinction between immediate operational performance and sustained data quality.
Participating sites
The registry assesses the number of sites able to participate and submit data for central analysis.
Data completeness
The registry assesses the completeness of submitted data.
Barriers to data submission
A survey is conducted post SPRINT-SARI study period on barriers to data completion.
Primary completion
The ClinicalTrials.gov record lists 2026-12 as the primary completion date.
This sequencing is statistically useful because early participation and later completeness represent related but distinct dimensions of data-collection performance. The study can therefore be viewed as addressing both the ability to establish data submission and the quality of the information ultimately submitted.
14. Limitations
- No posted statistical analyses: the ClinicalTrials.gov record does not provide formal statistical analyses for the three registered primary endpoints.
- Incomplete operational detail: the registry description of Data Completeness does not specify a numerical completeness formula in the information reported here.
- Undefined denominators: the registry defines the participating-site endpoint as a count but does not provide a denominator for converting that count into a proportion.
- Survey interpretation: barriers identified by a post-study survey describe reported barriers; they do not automatically establish causal relationships with incomplete submissions.
- No treatment-effect framework: the registry information does not identify randomized treatment groups or a therapeutic comparator, so conventional efficacy measures such as hazard ratios or risk ratios are not applicable to the primary endpoints as described.
- Completeness can depend on definition: element-level, record-level, and site-level completeness can produce different numerical summaries.
- Time-frame dependence: the one-week and 90-day endpoints measure different operational windows and should not be treated as interchangeable.
- Recruiting status: the registry lists the study as recruiting, with primary completion scheduled for 2026-12; the timing of any future reporting can therefore affect the availability of endpoint results.
15. What the Registry Can Establish Statistically
The registry record establishes a clear measurement framework even though it does not currently provide posted statistical analyses. The first endpoint is a site-level count, the second concerns data completeness over 90 days, and the third uses a post-study survey to characterize barriers to data completion.
The three endpoints should be analyzed according to their measurement scales rather than forced into a single treatment-effect framework. Counts, completeness measures, and survey responses each require an appropriate definition of the numerator, denominator, and observation period.
The phrase Data Completeness alone does not determine a unique statistical formula. A reproducible analysis requires the expected data elements or records to be defined and then compared with what was actually submitted.
SPRINT-SARI's primary statistical questions concern the reliability and completeness of clinical data collection. That makes endpoint definition, missing-data structure, denominators, and survey measurement central statistical issues even without a randomized efficacy comparison.
16. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The study defines three primary measurements with different scales and observation windows: a one-week site count, a 90-day data-completeness measure, and a 90-day post-study survey of barriers to data completion.
Operational interpretation
The endpoint structure evaluates whether sites can participate and submit centrally analyzable information, whether submitted data are complete, and which barriers are reported after the study period.
These two perspectives complement one another. Statistical interpretation determines how each endpoint should be quantified and summarized, while operational interpretation explains why those measurements are relevant to the performance of a multicenter clinical data-collection study.
17. Sources
- ClinicalTrials.gov: NCT02498587 — SPRINT-SARI.
Continue through the Clinical Biostats clinical-trial library
Explore additional clinical-trial statistical analyses and educational resources covering study design, endpoints, statistical methodology, and interpretation.
18. Record Summary
SPRINT-SARI provides a useful statistical example of a clinical study in which the primary questions concern data-collection performance rather than a randomized treatment effect. The registry identifies a planned enrollment of 15000 and three primary endpoints: the number of participating sites able to participate and submit data for central analysis over one week; Data Completeness over 90 days; and a post-study survey on barriers to data completion over 90 days.
The statistical interpretation of such a study depends heavily on precise endpoint definitions. A site count is directly summarized as a count, while completeness requires a transparent definition of expected and observed information. The barriers endpoint is naturally suited to descriptive survey summaries, with careful attention to response denominators and the possibility of multiple responses per participant.