This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Prometra's Utilization in Mitigating Pain (PUMP) (NCT00817596), a na na study. The primary outcome is Demonstrate That Prometra Programmable Pump System Accurately and Safely Delivers Medication in the Intrathecal Space, as Programmed..
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00817596 |
| Phase | NA |
| Condition | Intractable Pain |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 110 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Flowonix Medical |
| Results posted | Yes |
2. Clinical Question
Population
People with intractable pain were enrolled in a study of the Prometra programmable infusion pump. The registry lists an enrollment of 110. It does not report further eligibility details or participant characteristics in the information available for this analysis.
Intervention
The intervention was the Prometra infusion pump, a device intended to deliver programmed medication into the intrathecal space. The study description focuses on whether the pump accurately and safely delivered medication according to its programming.
Comparator
There was no separate comparator group. The study is recorded as a single-group study with no masking and no randomized allocation. Its results therefore cannot be read as a direct comparison between Prometra and another device, treatment, or usual care.
Primary question
The primary question was whether the volume of medication delivered over valid fill and refill sessions corresponded to the volume programmed for delivery. The registry defines accuracy using a per-participant delivered-to-programmed volume ratio, summarized as a mean and intended to be accompanied by a 90% confidence interval. The specified timeframe was six months, described as an acute study. Although the endpoint is defined, its observed mean and confidence interval are not reported in the registry record. The record therefore does not establish how close the observed delivery was to the programmed amount.
3. Trial Design
The PUMP study is registered as a completed, single-group device study of the Prometra programmable infusion pump in people with intractable pain. The sponsor was Flowonix Medical. The study began in 2007-01, and primary completion was recorded as 2008-03. Enrollment was 110. The listed phase and allocation are NA, and masking is listed as none. No randomized allocation or independent control group is described.
The study’s stated purpose was to evaluate safety and efficacy in accurately supplying medication to the intrathecal space for chronic pain treatment. The primary outcome was framed specifically around delivery accuracy: the volume measured as delivered was compared with the volume programmed to be delivered. The intended assessment timeframe was six months, described in the registry as an acute study. The outcome description specifies that valid scheduled and unscheduled fill or refill sessions contribute to a cumulative per-participant ratio.
The record identifies one primary outcome measure and no posted analyses. It does not report an observed endpoint mean, a confidence interval, or a statistical test. It also does not report a detailed statistical analysis plan, further inclusion and exclusion criteria, participant flow, or the number of participants with endpoint data. These absences limit interpretation of the completed study: the design and planned measurement can be described, but the primary delivery-accuracy result cannot be assessed from the registry information available.
Because the study had one group, any observed delivery ratio would describe performance within that group under the study’s measurement approach. It would not, by itself, estimate a comparative effect against another device or establish how outcomes would differ under another treatment strategy.
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.
Overall Study
| Stage | Prometra Intrathecal Pump System |
|---|---|
| STARTED | 110 |
| COMPLETED | 62 |
| NOT COMPLETED | 48 |
Analysis population for the primary outcome, as defined in the registry: Sample size was based on previous testing and data from similar devices. With a population of 60 subjects the 90% CI would be 95 ± 7, or 0.88 to 1.02. Therefore, approximately 60 patients with 6-months of data will provide adequate data to demonstrate accuracy, although up to 110 patients may be implanted to account for patient drop-outs.
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.
| Characteristic | Unit | Prometra Intrathecal Pump System |
|---|---|---|
| Participants analysed | 110 | |
| Age Continuous | years | 56 (SD 13) |
| Sex: Female, Male: Female | Participants | 51 |
| Sex: Female, Male: Male | Participants | 59 |
6. Primary Endpoint
The registry lists the primary outcome as Demonstrate That Prometra Programmable Pump System Accurately and Safely Delivers Medication in the Intrathecal Space, as Programmed., measured over 6 months - acute study.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | 90% Confidence Interval |
| Unit | % of programmed vol. actually delivered |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint was the mean percentage of programmed medication volume actually delivered, assessed over six months. The registry describes accuracy through a delivered-to-programmed volume ratio. For each participant, delivered volume from all valid scheduled and unscheduled fill or refill sessions was summed, as was the programmed volume from those sessions; the cumulative delivered amount was then divided by the cumulative programmed amount. The outcome is intended to reflect the relationship between the pump’s programmed delivery and the volume determined to have been delivered across those sessions, rather than a single refill measurement.
A mean summarizes the endpoint across participants, while a confidence interval describes the statistical uncertainty around that summary. The registry specifies a 90% confidence interval, but it does not report the observed mean or interval. As a result, the extent and precision of delivery accuracy in the enrolled group cannot be determined from the posted results.
In general, a delivery-volume ratio is useful for assessing whether actual delivery tracks programmed delivery. A ratio near the programmed amount would indicate close correspondence under the measurement definition, while departures could reflect under- or over-delivery. That general interpretation does not substitute for an observed estimate. The endpoint’s definition also depends on which fill and refill sessions were considered valid and on the volumetric determination of delivered drug volume; additional measurement details are not reported in the registry record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record gives a defined primary outcome but no observed primary outcome result. It specifies a mean delivered-to-programmed volume ratio and a 90% confidence interval for the six-month assessment, yet neither estimate appears in the posted results. Consequently, the central statistical question—how closely delivered medication volume corresponded to programmed volume, and with what precision—cannot be answered from this record.
The study enrolled 110 participants and used a single-group design. Enrollment describes the number entered into the study; it does not establish how many contributed usable six-month endpoint data. The number with endpoint observations, the number excluded from the analysis, and the handling of incomplete follow-up are not reported in the registry record. Those details matter because an endpoint summary based on a subset may not represent all enrolled participants in the same way.
The registry’s planning rationale refers to a population of 60 subjects and states that the 90% confidence interval would be 95 ± 7, or 0.88 to 1.02. It says that approximately 60 patients with six months of data would provide adequate data to demonstrate accuracy. This is a planning statement, not an observed outcome. It should not be read as the study’s measured mean, as evidence that the actual confidence interval had that width, or as a result showing delivery accuracy. The rationale says the sample size was based on previous testing and data from similar devices; further details of those inputs are not reported in the registry record.
The design also constrains interpretation. With no comparator group, the study can be understood as an evaluation of delivery and safety within the enrolled group, not as a randomized comparison. The record does not report a statistical analysis of the primary endpoint, subgroup estimates, or inferential results. No p-value or adjusted estimate is posted. The absence of these results is distinct from a finding of no effect: it means that the registry record does not give the estimates needed to judge the outcome.
Safety counts are reported separately. The record lists 11 participants with serious adverse events among 110 at risk and 110 with other adverse events among 110 at risk. These counts describe events recorded in the group; they do not establish causality or provide a comparison with another group. The deaths field is null, so a death count is not reported in the registry record. Serious adverse-event terms and their reported percentages should be interpreted as event listings, not as evidence that the device caused each event.
Overall, the registry establishes what the primary endpoint was intended to measure and provides some safety counts, but it does not provide the primary endpoint estimate required to judge the stated delivery-accuracy objective. The planned sample-size rationale cannot fill that gap.
9. Statistical Methods Explained
The primary outcome is a continuous measurement derived from volumes. The registry describes a ratio formed by dividing cumulative delivered drug volume by cumulative programmed drug volume for each participant. In general, a ratio expresses one quantity relative to another, making it possible to assess delivery in relation to the amount the pump was programmed to deliver. The record expresses the endpoint as the percentage of programmed volume actually delivered, but it does not post the observed mean.
What does the per-participant ratio represent?
The ratio combines all valid fill and refill sessions for a participant. Delivered volumes from scheduled and unscheduled sessions are accumulated for the numerator, and programmed volumes from valid sessions are accumulated for the denominator. This differs from treating every refill as an independent participant-level observation: the registry defines a cumulative participant-level measure. Sessions are repeated measurements within a participant, so they are not equivalent to separate people. The exact rules for declaring a session valid and the detailed volumetric measurement procedures are not reported in the registry record.
Using cumulative volumes means that the endpoint reflects the total delivery experience across the included sessions. In general, a ratio based on aggregated volumes is not necessarily the same as an unweighted average of session-by-session ratios. The registry’s description identifies the cumulative calculation, but it does not give individual observations or enough detail to reconstruct any participant’s value.
What would the mean and confidence interval convey?
The mean is a measure of the average endpoint across the participants included in an analysis. A confidence interval gives a range of values compatible with the data and the method used, and communicates uncertainty in the estimated mean. A narrower interval generally indicates greater precision, while a wider interval indicates less precision. The registry specifies a 90% confidence interval for the endpoint but does not report the interval or the mean. No conclusion about the observed accuracy or its precision can therefore be drawn from the endpoint definition alone.
In general, a confidence interval is not a guarantee that the true value lies within the particular interval, nor does an interval by itself establish whether a result meets a clinically meaningful criterion. Such interpretation would require a prespecified target or acceptance criterion and the observed estimate. The registry record does not report an acceptance threshold for the primary outcome. The sample-size rationale mentions 95 ± 7, or 0.88 to 1.02, as a projected interval for a population of 60 subjects; those figures are part of planning, not observed results.
Why does the single-group design matter?
A single-group study can describe outcomes among participants who received the device, but it does not supply a contemporaneous control group for comparison. In general, a control group can help separate changes associated with an intervention from changes that might occur for other reasons. Without one, a delivery-volume ratio can still describe delivery relative to programming, but it cannot estimate a comparative effect against another device or care approach. The record lists no randomized allocation and no masking.
The absence of posted analyses is also important. The record identifies one outcome measure and zero analyses. It does not report the estimator actually used for the mean, the method used to construct the confidence interval, assumptions about the distribution of ratios, or any adjustments. A familiar method should not be attributed to the study unless it is documented. In general, the analysis method should fit the outcome’s distribution and account for the study design; the registry information does not establish which method was used.
What is over-dispersion, in general?
Over-dispersion occurs when observed variability is greater than a statistical model assumes. It is a general concern for some data types and models because underestimated variability can make uncertainty appear too small. The primary endpoint is a participant-level volume ratio, and the record does not report a model or evidence of over-dispersion. It would therefore be inappropriate to claim that over-dispersion was present or that a particular adjustment was made. More generally, repeated measurements and variation between participants can influence uncertainty, which is why the analysis unit and the method for estimating precision should be clear.
The registry says the sample-size rationale drew on previous testing and data from similar devices and anticipated that approximately 60 participants with six-month data would provide adequate data for the accuracy objective. This is a design rationale, not evidence that the intended precision was achieved. The rationale does not replace the missing endpoint estimate, and no observed-power calculation can be inferred from it. Any sample-size planning illustration based on an observed effect would also depend on an observed effect estimate, which is not reported in the registry record.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists one primary outcome measure and zero analyses. The outcome is the mean percentage of programmed volume actually delivered, with a 90% confidence interval specified for the six-month timeframe. The record does not report a statistical analysis plan, any hypothesis-testing procedure, or adjustments for multiple comparisons. With one listed primary measure, the record identifies a principal endpoint, but it does not establish whether other outcomes or exploratory analyses were collected or how they were handled.
Missing-data handling is not reported in the registry record. The endpoint description includes valid scheduled and unscheduled fill or refill sessions, but it does not say how participants with incomplete follow-up, missing volume measurements, or invalid sessions were handled in the overall analysis. Nor does it report the number of participants with six-month data. In general, an analysis can be sensitive to why observations are missing; excluding incomplete cases, imputing values, or using other approaches can lead to different assumptions. No particular approach can be attributed to this study from the available information.
The registry also does not define the analysis population—for example, whether the primary summary was intended to include all enrolled participants or only those with evaluable delivery data. The enrollment of 110 should not be treated as the endpoint analysis denominator without a reported participant count for that analysis. These omissions limit assessment of selection and missing-data bias, as well as reproducibility. The outcome definition and planning rationale are available, but the analysis population, missing-data rules, and multiplicity procedures are not reported in the registry record.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Group 1 | 11 | 110 | 10.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Group 1 | 110 of 110 |
Most frequent serious adverse event terms
| Event term | Organ system | Group 1 |
|---|---|---|
| Implant Site Infection | Infections and infestations | 3 of 110 |
| Diplegia (temporary) | Nervous system disorders | 1 of 110 |
| Drug Withdrawal Syndrome | General disorders | 1 of 110 |
| Extradural abscess | Infections and infestations | 1 of 110 |
| Hematoma | Vascular disorders | 1 of 110 |
| Implant Site Cellulitis | Infections and infestations | 1 of 110 |
| Leukocytosis | Blood and lymphatic system disorders | 1 of 110 |
| Nausea | Gastrointestinal disorders | 1 of 110 |
| Post Lumbar Puncture Syndrome | Injury, poisoning and procedural complications | 1 of 110 |
| Procedural Pain | Injury, poisoning and procedural complications | 1 of 110 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Group 1 |
|---|---|---|
| Procedural Pain | Injury, poisoning and procedural complications | 37 of 110 |
| Implant Site Pain | General disorders | 24 of 110 |
| Nausea | Gastrointestinal disorders | 15 of 110 |
| Surgery (device-related) | Surgical and medical procedures | 13 of 110 |
| Implant Site Edema | General disorders | 11 of 110 |
| Implant Site Erythema | General disorders | 11 of 110 |
| Wound secretion | Injury, poisoning and procedural complications | 9 of 110 |
| Headache | Nervous system disorders | 9 of 110 |
| Vomiting | Gastrointestinal disorders | 8 of 110 |
| Post Lumbar Puncture Syndrome | Injury, poisoning and procedural complications | 8 of 110 |
Serious events with the highest arm-level rate: implant site infection (2.7%), diplegia (temporary) (0.9%), drug withdrawal syndrome (0.9%), extradural abscess (0.9%), hematoma (0.9%).
The registry lists 11 participants with serious adverse events among 110 at risk in Group 1. It also lists 110 participants with other adverse events among 110 at risk. These are group-level counts in the record; they do not identify how many events each participant experienced or establish whether an event was caused by the device. The deaths field is null, so the number of deaths is not reported in the registry record; null should not be interpreted as zero.
The most frequently listed serious adverse-event term is Implant Site Infection at 2.7%. Other listed serious terms are Diplegia (temporary), Drug Withdrawal Syndrome, Extradural abscess, Hematoma, and Implant Site Cellulitis, each at 0.9%. The registry gives these percentages as event-term figures. It does not provide further detail on timing, severity grading, relatedness, resolution, or whether participants could contribute more than one event term. They should therefore be read as descriptive safety listings rather than comparative risk estimates.
Because the study has a single group, the adverse-event counts cannot be contrasted with a control group within this study. They also do not, by themselves, establish that the device caused or did not cause an event. Interpretation is limited to the counts and terms recorded; additional safety-analysis details are not reported in the registry record.
12. What the Result Does, and Does Not, Mean
The study’s primary outcome was designed to assess how closely medication delivered by the Prometra pump matched the volume programmed for delivery over valid fill and refill sessions. The registry provides a detailed description of that cumulative per-participant measure and identifies a six-month assessment timeframe. It does not provide the observed mean or the 90% confidence interval. The principal accuracy result is therefore unavailable in the registry record, and the planned endpoint cannot be treated as a result.
The record does show that 110 participants were enrolled in a completed single-group study and gives group-level adverse-event counts. These registry entries describe the study’s size, design, and recorded safety information, but they do not substitute for the missing delivery-accuracy estimate. The planning rationale—based on previous testing and data from similar devices—describes the expected information from approximately 60 participants with six-month data. Its projected confidence interval is not an observed finding and cannot show what the study actually measured.
The single-group design also means that the record does not offer a randomized comparison with another device or treatment. Any endpoint estimate, if reported, would describe the enrolled group under the study’s measurement approach. It would not independently establish a difference from an alternative. Likewise, adverse-event counts do not establish causation, and the null deaths field does not establish that no deaths occurred.
In general, statistical interpretation depends on the observed estimate, its uncertainty, the analysis population, and the methods used to handle incomplete observations. Here, several of those elements are not reported in the registry record. The appropriate conclusion is limited: the study planned to measure cumulative delivery accuracy and recorded safety events, but the posted information does not permit assessment of the primary accuracy result or its precision.
13. Why This Trial Matters Statistically
This study is statistically notable because its primary endpoint is tied directly to a device’s programmed and measured output. Rather than relying only on a broad clinical outcome, the registry defines a cumulative delivered-to-programmed volume ratio across valid fill and refill sessions for each participant. That definition makes the measurement target explicit and distinguishes it from a single-session reading or an average of separate session ratios.
The study also illustrates why a clear endpoint definition is not enough to support a statistical conclusion. The registry states the planned summary—a mean with a 90% confidence interval—and provides a sample-size rationale based on earlier testing and similar devices. But the observed mean, interval, analysis population, and analysis details are not reported in the registry record. Without the estimate, the endpoint’s actual performance and precision cannot be evaluated.
The single-group structure is important as well. It permits description of the enrolled group’s measurements, but it does not offer a randomized control comparison. The endpoint may address delivery relative to programming, yet comparative claims would require evidence beyond this design and the posted information. The safety listings likewise describe recorded events without a control group or reported causal assessments.
For clinical researchers and statisticians, the record underscores the distinction between a protocol objective, a planning calculation, and an observed result. The projected 90% confidence interval in the sample-size rationale concerns anticipated precision, not achieved precision. The missing endpoint findings prevent readers from determining whether the planned information was obtained. Reporting participant-level analysis counts, missing-data methods, and the actual estimate with its uncertainty would be necessary for a more complete statistical interpretation.
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 statistical interpretation is that no primary endpoint estimate is reported. Although the registry defines a cumulative delivered-to-programmed volume ratio and specifies a mean with a 90% confidence interval, it does not give the observed mean or interval. The listed planning values cannot be substituted for those results.
The registry does not report how many of the 110 enrolled participants contributed six-month endpoint data, the analysis population, or the handling of participants with incomplete data. It also does not describe the rules for valid fill and refill sessions beyond including valid scheduled and unscheduled sessions in the calculation. These gaps make it difficult to assess completeness, selection, or reproducibility of the endpoint analysis.
The design is single group, with no randomized allocation and no masking. No comparator is identified, so the registry record cannot support a direct comparison with another device or care strategy. Further statistical methods—including the method for calculating the confidence interval, any model assumptions, and any sensitivity analyses—are not reported in the registry record. The record identifies one outcome measure and zero analyses.
Safety interpretation is also limited. The record lists serious and other adverse-event counts and several serious-event terms with percentages, but does not report event timing, relatedness, participant-level event patterns, or additional analysis details. The deaths field is null, and a death count is not reported in the registry record. These limitations mean the safety listings are descriptive and cannot establish causation or comparative risk.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Prometra's Utilization in Mitigating Pain (PUMP) — NCT00817596.
- Registry record last updated 2012-08-17; status last verified 2012-08.
- Results first posted 2012-07-31.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The PUMP study is a completed, single-group device study sponsored by Flowonix Medical, with enrollment of 110 and no masking. It evaluated the Prometra programmable infusion pump in people with intractable pain. The primary outcome was defined as the mean percentage of programmed medication volume actually delivered over six months, based on a cumulative per-participant ratio across valid scheduled and unscheduled fill and refill sessions.
The registry specifies a 90% confidence interval and gives a planning rationale based on prior testing and data from similar devices. It does not report the observed primary outcome estimate, the analysis population, or the statistical analysis details. Safety listings include 11 participants with serious adverse events among 110 at risk and 110 with other adverse events among 110 at risk. The deaths field is null, so a death count is not reported in the registry record.