Saxenda in Obesity Services (STRIVE Study): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Saxenda in Obesity Services (STRIVE Study) (NCT03036800), a randomized phase 4 study. The primary outcome is Weight Loss of ≥15% From Baseline (Complete Cases).

2. Clinical Question

The question concerns the result of adding a targeted medication pathway to the specialist service, rather than comparing medication with no care. Both groups received specialist obesity care; the pathway group also had targeted use of liraglutide under the stated stopping-rule approach. The trial was described as a real-world randomized study of otherwise largely unselected patients referred to specialist services.

The primary measure is a binary outcome: each included participant either met the weight-loss threshold or did not. The registry reports counts by group and a comparison between them. It also reports a logistic-regression analysis as an odds ratio. Those are different ways to express the group contrast, and the odds ratio should not be read as a ratio of participant counts or as a direct estimate of the difference in probabilities.

This is a question about the observed outcome in the trial’s defined analysis population. The record does not establish what the result would be in people outside the described service setting, or what would happen under a different prescribing pathway or follow-up arrangement.

3. Trial Design

STRIVE was a phase 4, randomized, parallel-group trial with an open-label design. The registry describes a two-year, real-world study of people referred to a Tier 3 or equivalent specialist weight management service for severe and complex obesity. Participants were assigned to standard care or to a targeted prescribing pathway that combined standard care with liraglutide 3 mg under prespecified stopping rules. The registry lists enrollment of 392 and marks the study as completed.

The primary outcome was assessed at 52 weeks. Its analysis population was not all enrolled participants: it was defined as randomized participants with data available for the outcome, analysed according to their baseline assignment, with participants who had bariatric surgery during the study period excluded. The reported group denominators for this outcome are therefore the complete-case denominators shown in the outcome table, rather than the overall enrollment figure.

Open-label means that participants and care teams were not masked to the assigned care pathway. The record does not report whether outcome assessment was masked. Because the outcome is based on a weight-loss threshold, the result is less dependent on a subjective rating than many outcomes, but an open-label design can still influence care, follow-up, and participant behaviour. This is a general design consideration, not evidence that such influence occurred.

The registry describes the study’s aim as comparing effectiveness, budget impact, and cost-effectiveness in a real-world setting. The results concern the posted primary weight-loss outcome and the additional outcomes included in the record. The registry record does not report the original sample-size justification or detailed protocol methods for allocation, concealment, or the statistical analysis plan.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Complete cases population The complete cases population is defined as all randomised participants who have data available for the outcome being analysed, according to the study group to which they were randomised at baseline. Participants who had bariatric surgery during the study period were excluded from this population.

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.

6. Primary Endpoint

The registry lists the primary outcome as Weight Loss of ≥15% From Baseline (Complete Cases), measured over 52 weeks.

The primary endpoint is the number of complete-case participants achieving weight loss of at least 15% from baseline at 52 weeks. It is a binary outcome: a participant is counted if the threshold is reached and is not counted otherwise. The registry reports 51 of 201 participants in the targeted prescribing pathway group and 6 of 93 in the standard-care group for this outcome.

A threshold endpoint gives a clear, clinically interpretable description of how many participants reached a specified level of weight loss. It also allows a direct comparison of the proportions reaching that level. However, converting weight change into a yes-or-no threshold discards information about how much weight participants lost below or above the threshold. Two participants with different amounts of weight loss can be classified identically if they fall on the same side of the threshold.

The complete-case definition matters to interpretation. Participants without available data for the outcome are not included in the stated analysis population, and those who had bariatric surgery during the study period were excluded. The registry record does not report the reasons outcome data were unavailable, or how missing outcomes were handled outside this definition. The result therefore describes the participants included in the specified analysis, not necessarily every randomized participant.

The 15% threshold is one of several weight-loss thresholds recorded in the trial. The primary question focuses on that threshold at 52 weeks; it should not be conflated with the additional thresholds reported at an earlier time point.

7. Results Posted in the Registry

• Standard Care Arm: 6.5% (6/93) versus 25.4% (51/201) on control; risk difference -18.9 percentage points (95% CI -26.7 to -11.1), risk ratio 0.25 (0.11–0.57); unadjusted two-proportion z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Weight Loss of ≥5% From Baseline (Complete Cases)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Standard Care Arm: 39.5% (34/86) versus 78.5% (186/237) on control; risk difference -38.9 percentage points (95% CI -50.5 to -27.4), risk ratio 0.50 (0.38–0.66); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Standard Care Arm than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Weight Loss of ≥10% From Baseline (Complete Cases)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Standard Care Arm: 10.5% (9/86) versus 25.3% (60/237) on control; risk difference -14.9 percentage points (95% CI -23.4 to -6.3), risk ratio 0.41 (0.21–0.80); unadjusted two-proportion z-test p = 0.004.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Standard Care Arm than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Weight Loss of ≥15% From Baseline (Complete Cases)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Standard Care Arm: 2.3% (2/86) versus 4.2% (10/237) on control; risk difference -1.9 percentage points (95% CI -6.0 to 2.2), risk ratio 0.55 (0.12–2.47); unadjusted two-proportion z-test p = 0.426.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Standard Care Arm and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a broad set of secondary outcome topics, including appointment attendance, treatment compliance, stopping treatment, and cessation due to adverse effects. It also reports weight-loss thresholds at 16 weeks. The displayed secondary results should be read as supporting context for the primary 52-week question, not as additional primary endpoints.

Across the posted 16-week weight-loss thresholds, the broad pattern is that the lower thresholds show clearer separation between the groups, while the highest threshold has a less precise comparison. This is consistent with the idea that the groups’ contrast can depend on the particular cutoff used. It does not establish a smooth or uniform pattern across all degrees of weight loss, because threshold outcomes classify participants into categories and omit the detail of continuous weight change.

The earlier assessment also differs from the primary endpoint in timing. It cannot answer the same question as the 52-week outcome, and the presence of multiple thresholds and time points creates more opportunities for apparent differences to arise by chance. The registry lists 81 outcome measures and 43 analyses; the multiplicity strategy, if any, is not reported in the registry record.

Unadjusted secondary p-values generally carry less weight than a prespecified primary comparison when many outcomes are examined. Without a reported multiplicity approach, they should be treated as exploratory evidence rather than as independent confirmation. The secondary outcomes may help describe the trial, but they do not replace the primary result or resolve limitations in the primary analysis population.

For the other listed secondary topics, the record’s titles identify what was assessed, but the information does not support a complete account of the direction or size of results across those topics. No conclusions about appointment attendance, treatment compliance, or treatment cessation should be inferred from the weight-loss analyses.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Participant Attendance at Appointments (Intention to Treat)
• Participant Attendance at Appointments (Intention to Treat)
• Cessation of Treatment Due to Adverse Effects (Intention to Treat)
• Cessation of Treatment Due to Adverse Effects (Intention to Treat)
• Participant Compliance With Treatment (Intention to Treat)
• Participant Compliance With Treatment (Intention to Treat)
• Patient Stopping of Treatment (Intention to Treat)
• Patient Stopping of Treatment (Intention to Treat)

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is binary because it records whether each participant crossed a specified weight-loss threshold by the stated time point. For a binary outcome, the simplest group summaries are the number meeting the definition and the number assessed in each group. The primary outcome table presents these counts for the targeted prescribing pathway and standard care. A proportion can also be expressed from those counts, but the comparison is most useful when its definition and direction are clear.

What does the risk difference describe?

The unadjusted recalculation reports a risk difference of -0.189, with a 95% confidence interval from -0.267 to -0.111. A risk difference is the difference between two event proportions. Here the comparison is oriented as standard care relative to the targeted pathway, so a negative value indicates that the observed event proportion was lower in standard care. The estimate is in probability-difference units; it is not an odds ratio. The confidence interval expresses uncertainty around this unadjusted estimate under the calculation used.

The recalculation is based on the posted summary counts and is unadjusted. It does not incorporate participant-level information or any covariates. It is useful as a transparent check on the basic group contrast, but it is not a replacement for the sponsor’s own analysis. No discrepancy is recorded between the recalculation and the sponsor analysis in the registry information.

Why is a logistic-regression odds ratio also reported?

Logistic regression is a method for modelling a binary outcome. Its effect estimate is commonly expressed as an odds ratio. The sponsor’s reported odds ratio for the primary outcome is 5.19, with a 95% confidence interval from 2.09 to 12.88 and p < 0.001. An odds ratio compares the odds of the outcome between groups. Odds and probabilities are related but are not interchangeable, so the odds ratio should not be interpreted as a risk ratio or as a direct probability difference.

The registry identifies the method as logistic regression but does not report the model covariates, the exact model specification, or whether the reported estimate is adjusted for particular baseline characteristics. It is therefore appropriate to describe the analysis as the reported regression analysis, not to assume what was included in it. The registry also does not report the prespecified statistical analysis plan.

How should a p-value be read?

A p-value describes how incompatible the observed data are with a stated null model, under the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of an effect. The primary unadjusted result is reported as p < 0.001, as is the sponsor’s primary logistic-regression result. The estimate and its confidence interval provide information about the direction and range of values compatible with the analysis; a p-value alone does not.

The secondary 16-week outcome at the 15% threshold has an unadjusted p-value of 0.426 and a reported regression p-value of 0.206. Its risk-difference confidence interval includes zero. This combination indicates that the posted analysis does not give clear evidence of a difference for that secondary comparison. It does not prove that the groups are identical.

Why can over-dispersion matter?

Over-dispersion is a general concern when data vary more than a statistical model expects. It can make standard errors too small if the model does not allow for the extra variation, leading to confidence intervals that are too narrow and p-values that are too small. It is especially relevant to some count and clustered-data analyses. The primary outcome is a binary participant outcome, and the registry identifies logistic regression for the sponsor analysis. The registry record does not report a specific over-dispersion assessment or correction, so no claim can be made about whether it was considered.

More generally, methods depend on how observations were collected and how the outcome is distributed. Random assignment supports a comparison of assigned groups, but it does not remove uncertainty from missing outcomes, exclusions, or the choice of analysis population. The complete-case definition and exclusion of participants who had bariatric surgery are therefore part of the interpretation, not incidental details.

11. Confidence Intervals and Precision

The primary unadjusted risk-difference estimate is -0.189, with a 95% confidence interval from -0.267 to -0.111. This interval gives a range of values compatible with the observed data under the calculation and its assumptions. It indicates that the estimated difference is not highly precise at a single value: the interval spans a range of possible risk differences. The entire interval is below zero under the comparison’s standard-care-minus-pathway direction.

A confidence interval is not a guarantee that the true effect lies inside its endpoints, nor does it describe the range of individual participant outcomes. In general, a narrower interval indicates greater statistical precision than a wider one, while the interval’s position relative to zero helps describe whether the data are compatible with no risk difference under the method used. Interpretation should consider both the estimate and the interval, rather than relying only on whether a threshold for statistical significance is crossed.

The sponsor’s primary logistic-regression estimate is an odds ratio of 5.19, with a 95% confidence interval from 2.09 to 12.88. This interval is on the odds-ratio scale, so it cannot be compared directly with the risk-difference interval by looking at their numerical widths. The two intervals describe different effect measures. The registry record does not provide details of the regression model sufficient to explain why its estimate differs in form from the unadjusted calculation.

Precision is also influenced by the number of observations contributing to the analysis and by the number of outcome events. The primary complete-case denominators are 201 and 93, with 51 and 6 participants meeting the threshold, respectively. The unequal denominators and the relatively small event count in standard care are relevant context for reading the interval, but do not by themselves establish a particular source of uncertainty. Reasons for unavailable outcome data are not reported in the registry record.

The 16-week secondary outcome at the 15% threshold has an unadjusted risk-difference interval from -0.060 to 0.022. Because this interval includes zero, the result is compatible with no difference as well as with values on either side under that calculation. Its uncertainty should be distinguished from the more clearly separated primary 52-week result.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size table is a planning illustration based on the observed effect, not the original trial design calculation. Its figures are participants per arm, rather than total events. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect.

The 80% power column shows the estimated participants per arm for a future study to have an 80% chance of detecting the specified effect, if that effect and the planning assumptions are correct. The 90% power column describes the corresponding planning target for a 90% chance. These are design calculations for hypothetical future studies; they do not describe the probability that the completed trial’s result is correct.

Rows that plan for 100%, 75%, or 50% of the observed effect illustrate how the required sample changes when the assumed effect changes. A smaller effect is harder to distinguish from random variation, so detecting it generally requires more participants. The table’s values should be read as conditional planning quantities, not as evidence that a future trial will reproduce the observed effect.

This is not observed power. The registry record does not report the assumptions used in the original design or its sample-size justification. The planning illustration on this page uses the stated assumptions and the observed effect as a reference, and it should not be mistaken for the trial’s prospective calculation or for an assessment of whether the completed study was adequately powered.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.254; treatment proportion p₁ = 0.065; average p̄ = 0.159.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5173 = 1.0139.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4997 = 0.4206.
4. n = (1.0139 + 0.4206)² / (-0.1892)² = 57.5, rounded up to 58 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.253731343283582                          # control proportion
p1 <- 0.0645161290322581                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3036800)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 58 (80% power), 77 (90% power)
 75% of observed effect: n per arm = 116 (80% power), 154 (90% power)
 50% of observed effect: n per arm = 287 (80% power), 383 (90% power)
power.prop.test: 57.5 
Simulated power at n = 58 per arm: 0.825 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry’s sponsor-reported primary analysis uses logistic regression and reports an odds ratio of 5.19, with a 95% confidence interval from 2.09 to 12.88 and p < 0.001. The recalculation shown alongside it is an unadjusted risk difference based on the posted summary counts: -0.189, with a 95% confidence interval from -0.267 to -0.111 and p < 0.001.

These estimates are not expected to be numerically identical because they use different effect scales and different methods. The risk difference expresses an absolute contrast in event proportions, while the odds ratio compares odds. The registry identifies logistic regression but does not report enough model detail to say which covariates, if any, were included. The recalculation uses only the posted group counts and is not an adjusted analysis.

The registry information records no discrepancy between the recalculation and the sponsor’s analysis. The recalculation is therefore best read as a transparent, complementary summary of the raw group contrast, while the sponsor’s own regression analysis remains the reported trial analysis. Neither estimate should be substituted for the other or interpreted without its scale, analysis population, and uncertainty interval.

15. Multiplicity, Missing Data and Analysis Populations

The record includes many outcomes and analyses, making multiplicity relevant to interpretation. When investigators test multiple outcomes, thresholds, or time points, the chance of at least one apparently unusual result can increase. This is a general statistical concern; it does not show that any particular finding in this trial is spurious. The registry lists 81 outcome measures and 43 analyses, but a plan for controlling the impact of multiple comparisons is not reported in the registry record.

The primary endpoint is identified as weight loss of at least 15% from baseline at 52 weeks. The record also lists secondary weight-loss outcomes at 16 weeks and other topics, including attendance, compliance, treatment stopping, and cessation due to adverse effects. Results from secondary analyses should therefore be interpreted with attention to their status and the number of questions examined. Whether analyses were prespecified, whether a hierarchy was used, and whether any adjustment was applied are not reported in the registry record.

The primary analysis population is defined as randomized participants with data available for the outcome, analysed in their assigned groups, excluding participants who had bariatric surgery during the study period. This complete-case approach makes the available outcome data central to interpretation. The registry record does not report how many participants lacked primary outcome data for each reason, why data were unavailable, or whether missing outcomes were imputed or otherwise handled in another analysis. The record also does not provide a sensitivity analysis for missing outcomes in the information.

The overall enrollment is 392, while the primary outcome table gives complete-case denominators of 201 for the targeted pathway and 93 for standard care. These figures describe different quantities: overall enrollment and participants contributing to the stated primary outcome analysis. The registry’s definition explains that outcome availability and the surgery exclusion affect inclusion, but does not give a full accounting of the difference.

Randomized assignment and analysis according to assigned group are important features of the stated population definition. However, the record does not report the original allocation-concealment procedures, the detailed statistical analysis plan, or further analysis-population definitions for every outcome. These omissions limit assessment of how all analysis choices were made.

16. 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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: abdominal infection (2.3%), urinary tract infection (1.5%), cholecystitis (1.2%), viral infection (1.2%), pancreatitis (0.8%).

The registry reports serious adverse-event figures for the two groups. In the targeted prescribing pathway group, 43 of 260 participants were reported as affected by a serious adverse event; in standard care, the corresponding figures were 16 of 132. The record also reports other adverse-event figures: 238 of 260 in the pathway group and 89 of 132 in standard care. It records 2 deaths in the pathway group and 0 in standard care.

These are descriptive figures from the registry record. They do not, by themselves, establish that the assigned care pathway caused an event or that an event was related to liraglutide. The record information does not report event-level timing, severity details beyond the serious classification, or attribution to treatment. The denominators shown for safety also differ from the complete-case denominators used for the primary outcome, and safety counts should not be substituted for outcome-analysis counts.

The registry lists serious-event terms including abdominal infection, urinary tract infection, cholecystitis, viral infection, pancreatitis, and bronchospasm, with the percentages shown in the safety table. Such listings describe reported event categories; they do not show whether events were judged treatment-related. The absence of a reported event in a category is not evidence that it could not occur.

These data support a descriptive account of events recorded during the study, not a standalone comparative safety conclusion. The registry record does not provide the additional detail needed to assess event relatedness or explain differences in exposure and follow-up.

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

The primary result describes the proportion of complete-case participants in each assigned group who reached a defined weight-loss threshold at 52 weeks. In the posted data, more participants in the targeted prescribing pathway group than in standard care met the threshold, and both the unadjusted comparison and the sponsor’s regression analysis report evidence of a group contrast. The estimates and confidence intervals convey the size and uncertainty of that contrast using different statistical scales.

This result does not mean that every participant in the pathway group reached the threshold, or that no participant in standard care did. It does not describe the amount of weight change for people who did not reach the cutoff. Nor does it establish what would happen for people who were not referred to a Tier 3 or equivalent specialist service, or under a different medication protocol, stopping rule, or care setting.

The comparison is between two service strategies: specialist care alone and specialist care with targeted use of liraglutide 3 mg. It does not isolate the medication’s contribution from other aspects of the pathway, such as monitoring or changes in clinical contact. The registry record does not report information that would allow those components to be separated.

The primary analysis is complete-case and excludes participants who had bariatric surgery during the study period. Therefore, the estimate applies directly to the analysis population as defined, and missing outcome data may affect how representative that population is of all randomized participants. The reasons for missingness and results under alternative missing-data assumptions are not reported in the information.

The finding is statistical evidence from a randomized trial, not a guarantee of an identical result in other populations or services. It does not establish the balance of benefits, harms, costs, or individual preferences. The registry describes broader effectiveness and economic aims, but the information does not support conclusions about budget impact or cost-effectiveness.

18. Why This Trial Matters Statistically

This trial is statistically informative because it evaluates a care pathway in a specialist-service setting rather than considering medication use in isolation. Random assignment provides a framework for comparing the two assigned strategies, while the open-label design reflects the practical nature of the intervention and leaves some aspects of care potentially visible to participants and staff.

The primary outcome is a threshold-based binary measure, and the registry presents both group counts and a regression estimate. Seeing these alongside an unadjusted recalculation helps distinguish absolute differences in event proportions from relative comparisons of odds. That distinction matters: a single headline estimate can obscure the scale on which a result is expressed.

The trial also illustrates why analysis population definitions matter. Overall enrollment is not the same as the number contributing to a complete-case endpoint. Here, the primary population excludes participants without available outcome data and those who had bariatric surgery during the study period. Clear reporting of those decisions helps readers understand exactly which participants the estimate describes, even though the record does not give a full account of missing data.

Finally, the record includes numerous outcome measures and analyses. This makes it important to distinguish the primary result from secondary findings, especially when multiplicity handling is not reported. The study is a useful example of reading a registry result with attention to endpoint definition, effect scale, uncertainty, analysis population, and the limits of the posted methods.

19. 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.

Several limitations specific to this trial shape interpretation. It was open-label, so participants and care teams knew the assigned pathway. The registry record does not say whether outcome assessment was masked. In a service-based intervention, awareness of assignment could influence aspects of care or follow-up; that is a general possibility, not evidence that it occurred in this study.

The primary result is based on a complete-case population rather than all enrolled participants. The definition excludes participants who had bariatric surgery during the study period and requires outcome data to be available. The record does not report the reasons outcome data were unavailable, the distribution of missingness by reason, or sensitivity analyses under alternative assumptions. This limits assessment of how missing outcomes might affect the estimate.

The reported group denominators for the primary endpoint are unequal, and the event count in the standard-care group is small relative to the pathway group. The confidence interval provides a way to see the resulting uncertainty, but the registry does not provide further detail about the reasons for the available denominators. The sponsor’s regression model is identified as logistic regression, but its covariates and specification are not reported.

The setting is also specific: participants had severe and complex obesity and were referred to Tier 3 or equivalent specialist services. The posted result should not automatically be generalized to other populations or service arrangements. The intervention was a pathway combining specialist care with targeted liraglutide use and stopping rules, so the result does not isolate the contribution of each pathway component.

Finally, although the study’s stated aim included budget impact and cost-effectiveness, the information does not provide those results. The registry record also does not report the original sample-size justification, detailed statistical analysis plan, or multiplicity strategy. These omissions limit assessment of design assumptions and the interpretation of the full set of outcomes.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Saxenda in Obesity Services (STRIVE Study) — NCT03036800.
• Registry record last updated 2024-08-13; status last verified 2023-01.
• Results first posted 2024-08-13.
• 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.

23. Record Summary

STRIVE (NCT03036800) is a completed phase 4, randomized, parallel-group, open-label study of a targeted prescribing pathway for people referred to specialist obesity services. The pathway combined standard care with targeted liraglutide 3 mg under prespecified stopping rules; the comparator was standard Tier 3 specialist care alone. The registry lists enrollment of 392.

The primary outcome was the number of complete-case participants achieving weight loss of at least 15% from baseline at 52 weeks. The posted counts were 51 of 201 in the pathway group and 6 of 93 in standard care. The sponsor’s logistic-regression analysis reports an odds ratio of 5.19, with a 95% confidence interval from 2.09 to 12.88 and p < 0.001. An unadjusted recalculation reports a risk difference of -0.189, with a 95% confidence interval from -0.267 to -0.111 and p < 0.001.

Interpretation is specific to the defined complete-case population, the specialist-service context, and the methods reported in the registry. Detailed missing-data handling, multiplicity strategy, original sample-size justification, and economic results are not reported in the registry record.