MICRO Study: Detecting Bacterial Infections Related to Orthopaedic Surgical Implants: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for MICRO Study: Detecting Bacterial Infections Related to Orthopaedic Surgical Implants (NCT03132246), a registered phase not stated study. The primary outcome is Number of Blood Samples With High Levels of Absorbance Via ELISA for Patients (Greater Than 1.5 Ratio Compared to Control Human Serum).

2. Clinical Question

3. Trial Design

The MICRO Study, registered as NCT03132246, is described as a prospective trial of blood testing in trauma patients with fractures treated using metal implants. The registry record lists Open Reduction and Internal Fixation of Extremity Fractures as the condition. The sponsor is the University of Maryland, Baltimore. The study began in 2014-04, had a primary completion date of 2018-06-30, and is listed as COMPLETED. The registry enrollment figure is 260.

The study description says patients were to have between 1-3 routine blood draws for laboratory analysis. Patients could be infected or not infected when blood was drawn, and selection for the study was described as random. That statement concerns selection of patients, not random assignment to a treatment. The record does not state that participants were assigned to interventions, and it explicitly cautions that the infected and not infected groups are not a like-for-like control comparison.

The arm descriptions characterize both groups as trauma patients considered at risk of infection after fracture fixation. They list open fractures, proximal tibia fractures, pilon fractures, and calcaneus fractures treated with surgical fixation as inclusion criteria. The registry says participants would be followed by their treating physicians according to standard care and that clinical cultures would be monitored during the one-year post-operative period.

The listed intervention is blood draws, classified as a diagnostic test. The primary outcome is a count of blood samples with high ELISA absorbance; a secondary outcome concerns samples with low absorbance. The registry record reports 2 outcome measures and 0 analyses. It does not report a statistical analysis plan or explain how repeated samples from the same participant were handled.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: Patients were enrolled if presented as high risk but later separated into the Infected Group or the Not Infected Group based on whether or not they went on to develop an infection post surgery. Group assignment is unknown for patients that did not complete the study and therefore groups are combined for reporting overall study numbers.

Overall Study

Analysis population for the primary outcome, as defined in the registry: One to three blood sample tests were taken for each participant prior to the one-year post-operative clinical follow-up. Clinical cultures will also be monitored for the one-year post-operative period and compared to the blood sample test results to determine if the test is able to identify patients with positive staphylococcal infection cultures.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: Trauma patients that were considered at risk for infection after fracture fixation. Inclusion criteria included open fractures, proximal tibia fractures, pilon fractures, and calcaneus fractures treated with surgical fixation.

6. Primary Endpoint

The registry lists the primary outcome as Number of Blood Samples With High Levels of Absorbance Via ELISA for Patients (Greater Than 1.5 Ratio Compared to Control Human Serum), measured over 1 year of clinical follow-up.

The primary endpoint is the number of blood draw samples with high levels of absorbance via ELISA, using the registry’s stated threshold of greater than 1.5 times the control human serum ratio. A high reading is defined in the record as a positive test result. The endpoint’s unit is blood draw samples, not participants.

The stated reason for counting positive samples among patients classified as not infected is to demonstrate the test’s false-positive rate. The registry description also says clinical cultures would be monitored during the one-year post-operative period and compared with blood test results to assess whether the test identifies patients. The endpoint therefore concerns diagnostic classification in relation to infection status, rather than the effect of a treatment.

Because each participant could provide 1-3 samples, a sample count may include repeated measurements from the same person. The record does not report whether the endpoint was also summarized per participant, how infection status was determined for each sample, or how samples and cultures were matched over time. Most importantly, no primary endpoint analysis is reported in the registry record. The endpoint definition explains what was intended to be counted, but it does not provide an observed count or a measure of diagnostic accuracy.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods determine how observed measurements are summarized, compared, and interpreted. The registry record identifies the planned outcome measures and their units, but it reports 0 analyses. It does not report the analytic procedures used for the outcomes, estimates, confidence intervals, or hypothesis tests. The concepts below are general explanations of issues that matter for this study; they should not be mistaken for methods documented in the registry record.

What does a count endpoint represent?

A count endpoint records how many units meet a stated definition. Here, the unit is a blood draw sample, and the event of interest is a high ELISA absorbance result under the registry’s threshold definition. A count is not itself a rate. To interpret a count as a proportion or rate, a denominator and a clearly defined set of eligible observations are needed. The record does not report primary outcome counts or the corresponding denominators.

Counts can be analyzed in different ways depending on the design and the question. A simple count may be described directly. A rate requires an exposure basis or denominator, while a proportion requires a defined number of opportunities for the event. Models for counts may account for differences in observation time or other factors, if those are part of the study’s analysis. The registry record does not say that any particular count model or rate calculation was used.

Why do repeated samples matter?

Participants could provide 1-3 blood samples. Measurements from the same participant are generally more alike than measurements from unrelated participants because they share characteristics and clinical history. In general, treating correlated observations as independent can make statistical uncertainty appear smaller than it is. Common approaches can account for within-participant clustering or summarize observations at the participant level, but the record does not report which approach, if any, was used.

The distinction between sample-level and participant-level questions also matters for clinical interpretation. A sample-level count answers how many tested samples met the threshold. A participant-level summary might instead ask how many people had at least one positive sample, or how a person’s results changed over time. Those are different estimands and can yield different summaries. The registry defines the primary endpoint in samples and does not report a participant-level primary analysis.

What does a confidence interval communicate?

In general, a confidence interval describes the statistical uncertainty around an estimate under the assumptions of the analysis. It is not a guarantee that the true value lies within a particular interval, and its width depends on factors including the amount of information and the variability in the data. For a diagnostic measure, an interval helps readers assess precision as well as the point estimate. No confidence interval for the primary endpoint or a diagnostic accuracy measure is reported in the registry record.

What do diagnostic accuracy measures require?

Diagnostic accuracy summaries compare test classifications with a reference classification, such as infection status established through clinical follow-up. Sensitivity concerns the proportion of infected cases that test positive; specificity concerns the proportion of non-infected cases that test negative. A false-positive rate concerns positive tests among those classified as not infected. These measures require a defined classification for each observation and the relevant positive and negative counts. The record describes clinical culture monitoring and a false-positive-rate rationale, but does not report the cross-classified results needed to calculate these measures.

Predictive values address the probability of a condition given a test result and depend on the distribution of the condition in the population being evaluated. They are not interchangeable with sensitivity or specificity. The registry record does not report predictive values or the outcome data needed to estimate them. Since patient selection and the represented clinical population affect interpretation, these quantities would also need context beyond the threshold definition alone.

Why can over-dispersion matter?

In general, a count model can fit poorly when observed variation exceeds what the model expects. This is called over-dispersion. It can arise when observations are clustered, when risks vary across participants, or when the model leaves important structure unaccounted for. If over-dispersion is present and ignored, standard errors and confidence intervals may be too optimistic. Repeated samples per participant are one reason a count analysis might need to consider dependence or extra variation. The record does not report a dispersion estimate, a model, or a check for over-dispersion.

What can the planning illustration establish?

The registry materials describe sample-size planning illustrations based on a stated fraction of an observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. These are planning assumptions, not results from the trial and not observed power. The record does not report the resulting participant-per-arm figures in the registry record for interpretation here, nor does it report a trial-specific sample-size justification. The allocation assumption in a planning illustration should not be confused with the study’s design: the registry says the groups are not a like-for-like control comparison and does not state an allocation method.

Because no outcome analyses are posted, there is no estimate to evaluate for magnitude or precision. A statistical method cannot replace the missing outcome counts, denominators, and uncertainty estimates. The record therefore permits explanation of the endpoint and general analytic considerations, but not an independent assessment of the test’s performance.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists 2 outcome measures: the primary count of samples with high ELISA absorbance and a secondary count of samples with low absorbance. It reports 0 analyses. No multiplicity procedure, hierarchy for interpreting the outcomes, or adjustment for testing more than one outcome is reported in the registry record. In general, evaluating multiple outcomes can increase the chance of at least one apparently unusual result when each is tested separately. Whether any adjustment is needed depends on the scientific questions and analysis plan, which are not reported.

The record does not report how missing blood samples, unavailable laboratory results, or incomplete clinical culture follow-up were handled. It also does not describe whether missingness was summarized by infection-status group or whether any imputation or sensitivity analysis was performed. These details matter because a count based only on available samples may not represent all planned testing if availability differs across participants or groups. This is a general concern, not evidence that missing data occurred.

The analysis population is not defined in the registry outcome information. The record does not state which enrolled participants were included in each outcome analysis, how the infected and not infected classifications were assigned, or whether any participants were excluded from outcome summaries. It also does not report whether outcome analyses used all available samples or a prespecified subset. With no analyses reported, there is no basis for assessing the impact of exclusions, missing observations, or alternative population definitions.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

The adverse-event listing reports 0 serious events and 0 other events in each of the listed groups. The groups are titled Unmasked Non-completers, Infected, and Not Infected; the corresponding at-risk figures are 104, 20, and 136. The registry also reports deaths of 0, 1, and 3 in those groups, respectively. These are descriptive registry figures and should be read alongside the group labels and denominators rather than as a treatment comparison.

The record does not report a causal assessment linking deaths to the blood draws, ELISA testing, infection status, or any intervention. The study is not described as a randomized treatment comparison, and the listed intervention is diagnostic blood draws. The figures therefore do not establish that the study procedures caused or prevented an event. The registry’s empty list of top serious-event terms is not an explanation of the deaths or of how events were reviewed.

Zero reported serious or other events in a group does not establish that the underlying risk is zero. It describes the entries in the registry record, not a complete assessment of all possible harms. No additional detail about event definitions, event timing, severity grading, or attribution is reported in the registry record.

12. What the Result Does, and Does Not, Mean

The registry record describes a planned way to evaluate whether ELISA absorbance in blood samples is associated with infection status after fracture fixation. It defines a high absorbance result as positive and explains why positive results among patients classified as not infected are relevant to false positives. The record also describes clinical culture monitoring during the one-year post-operative period.

However, the posted results do not include analyses for either outcome measure. There is no reported primary outcome count, no comparison of high-absorbance findings by infection status, and no uncertainty measure. The record therefore does not establish the frequency of positive test results, the false-positive rate, or the ability of the test to identify infection. It also does not provide evidence about the clinical consequences of using the test.

The distinction between a diagnostic study and a treatment trial is important. The listed intervention is blood draws for laboratory analysis, and the registry does not describe random assignment to a treatment. The infected and not infected groups are explicitly not characterized as like-for-like controls. Any observed contrast between such groups would require careful interpretation and would not, by itself, demonstrate a causal treatment effect.

The reported enrollment figure and completed status do not substitute for outcome data. Nor can a planning illustration establish the precision or performance of a test when the corresponding results are not reported. The registry record supports understanding the study question, population, sampling approach, and intended follow-up; it does not support a conclusion about whether the test did or did not meet its diagnostic aims.

13. Why This Trial Matters Statistically

This study raises statistical issues that recur in diagnostic research: defining what counts as a positive result, choosing the unit of analysis, comparing results with a reference classification, and reporting uncertainty. The primary outcome is expressed as a number of blood samples, while participants could contribute 1-3 samples. Clear handling of repeated observations is important because multiple samples from one person are not necessarily independent.

The registry also makes the distinction between test positivity and diagnostic accuracy visible. A threshold can define a positive result, but it cannot by itself tell readers how often the test identifies infection or produces a positive result among people classified as not infected. Those questions require the observed results and denominators, ideally with uncertainty estimates and clear definitions of infection status.

The record’s warning that the infected and not infected groups are not a like-for-like control comparison is consequential. Diagnostic group contrasts can be influenced by differences in who enters each group and how infection status is established. Without a reported analysis plan and outcome results, readers cannot judge those influences or assess precision. This is an example of why a study’s design, endpoint, and analysis population need to be interpreted together.

Finally, completed status and an enrollment figure do not answer the statistical question. Transparent reporting of outcome counts, denominators, methods for repeated measurements, missingness, and uncertainty would be needed to evaluate the stated diagnostic aim. The registry record documents the question and approach, but reports no outcome analyses.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The main limitation for interpreting this registry record is that it reports 0 analyses for the 2 outcome measures. There are no posted outcome counts or estimates from which to assess test performance. The endpoint definition and follow-up plan do not reveal what results were observed.

The endpoint counts blood samples, although each participant could contribute 1-3 samples. The record does not describe whether the analysis accounted for repeated measurements within a participant or whether a separate participant-level summary was planned. It also does not explain how a sample’s result was linked to infection status or clinical cultures over the one-year follow-up.

The infected and not infected groups are not a randomized or like-for-like control comparison according to the registry record. The record does not provide additional details about group classification, timing of infection, or the reference standard used to determine infection status. Consequently, even if group counts were available, interpretation would depend on information not reported in the registry record.

The statistical analysis plan, approach to missing samples or cultures, analysis populations, and multiplicity handling are not reported. The registry also does not provide confidence intervals or other measures of uncertainty for the diagnostic endpoint. These omissions prevent assessment of the reliability and precision of any apparent differences. They do not show that a particular problem occurred; they mean that the registry record does not report enough detail to evaluate it.

The safety listing is also limited: it provides group-level event figures without reported attribution, timing, or further explanation. These data should not be treated as evidence that the blood draws caused or did not cause an event.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: MICRO Study: Detecting Bacterial Infections Related to Orthopaedic Surgical Implants — NCT03132246.
• Registry record last updated 2020-05-15; status last verified 2020-05.
• Results first posted 2020-05-15.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

NCT03132246, the MICRO Study, is a prospective diagnostic-testing study of blood samples from trauma patients with fractures treated using metal implants. The registry lists the study as COMPLETED, with enrollment of 260, a start date of 2014-04, and primary completion on 2018-06-30. The listed intervention is blood draws for ELISA laboratory analysis.

The primary endpoint is the number of samples with high absorbance under the registry’s stated threshold; the record describes positive results among patients classified as not infected as relevant to the false-positive rate. Clinical cultures were to be monitored during the one-year post-operative period. The registry lists 2 outcome measures and 0 analyses. It reports no outcome estimates or uncertainty intervals, so the diagnostic performance cannot be determined from the registry record.