Implementation of Knowledge-Based Palliative Care: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Implementation of Knowledge-Based Palliative Care (NCT02708498), a non-randomized na study. The primary outcome is World Health Organization Quality of Life-BREF (WHOQOL-BREF).

2. Clinical Question

The project also aimed to examine staff implementation and leadership, and to identify barriers and facilitators to implementation. The registry lists staff versions of the Person-centred Care Assessment Tool and the Person-Centred Climate Questionnaire among secondary outcomes. These aims span both patient-reported outcomes and implementation questions; they are not all answered by the primary quality-of-life measure.

3. Trial Design

This was a completed, nonrandomized study with a sequential model and no masking, according to the registry record. The listed enrollment was 1151. The sponsor was Lund University. The study start date was 2015-04-09, and the primary completion date was 2017-06-20. The registry identifies the condition as ageing and gives the phase as NA.

The intervention was educational and behavioral. It was delivered in nursing homes in Kronoberg and Skåne. In each region, the registry describes intervention homes and control homes; control homes received no intervention. The registry description says the educational intervention was provided to ten nursing homes in each region and that the control group consisted of an equal number of nursing homes. Nonrandom allocation means that assignment did not use randomization. Differences between groups may therefore reflect pre-existing differences between homes, residents, staff, or regions as well as any effect associated with the educational program.

The project included older people, next of kin, staff, managers, and staff focus groups in its listed groups and safety record. These groups have different roles and different outcome information; their counts should not be combined as if they were a single outcome population. The listed inclusion criteria for older people required age 65 years or older, Swedish language, no dementia, and enough energy for a structured interview of up to one hour.

The registry record does not report the sequence used to introduce the program, the timing of measurements beyond the primary outcome timeframe, or details of how homes were allocated. It does not report a statistical analysis plan or an adjusted comparison. These omissions limit assessment of how the design handled the nonrandomized allocation and what differences between groups were considered in analysis.

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.

Only the first 6 of 8 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The inclusion criteria for the older persons were that they were 65 years or older, spoke Swedish, did not have dementia and had enough energy to participate in a structured interview for up to one hour.

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.

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as World Health Organization Quality of Life-BREF (WHOQOL-BREF), measured over 9 months.

The primary endpoint was the World Health Organization Quality of Life-BREF (WHOQOL-BREF), recorded as a score on a scale at 9 months. The registry describes the measure as assessing quality of life for older people living in nursing homes at the end of life. It specifies a five-point Likert scale and states that higher values mean better outcomes. The total score ranges from 26 to 130.

The registry also gives ranges for subscales: Overall Quality of Life and General health each range from 1 to 5; Physical Health ranges from 7 to 35; Psychological ranges from 6 to 30; Social relationships ranges from 3 to 15; and Environment ranges from 8 to 40. These dimensions allow the measure to represent aspects of quality of life rather than a single narrow clinical event. A score is an outcome measure, not by itself an explanation of why participants may report better or worse quality of life.

The registry specifies a mean as the summary parameter and an inter-quartile range as the dispersion measure. In general, a mean describes the average score, while an inter-quartile range describes the spread of the middle portion of observed scores. The registry record does not report the observed WHOQOL-BREF scores, the number contributing data at 9 months, or a comparison between intervention and control groups. Thus, the endpoint's scale and intended interpretation are specified, but its observed results are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a nonrandomized, sequential study with no masking. These design features affect how a comparison should be understood. Nonrandomized allocation means that intervention status was not assigned by chance. In general, randomization helps balance both measured and unmeasured characteristics across groups in expectation. Without it, differences in outcome may reflect differences in the homes or people as well as the intervention. The registry record does not explain the allocation sequence or report an adjustment strategy.

What does the sequential model tell us?

The registry labels the model sequential, but does not report the operational details of the sequence. The label alone does not establish when or how intervention homes began the program, whether data collection periods overlapped, or how time-related changes were handled. Changes in staffing, care practices, or the surrounding setting could potentially coincide with implementation. Whether such changes occurred or were considered in analysis is not reported in the registry record.

How should a nonrandomized comparison be read?

In general, an unadjusted group comparison describes observed differences between the groups as they were recorded. It does not necessarily estimate the causal effect of the intervention. Adjustment aims to account for measured factors that could be associated with both group assignment and outcome. Adjustment cannot automatically remove bias from factors that were not measured, measured poorly, or handled incorrectly. The registry record does not report a model, covariates, assumptions, or adjusted results, so the reader cannot evaluate these aspects of the analysis.

What do the endpoint summaries mean?

The registry specifies a mean and an inter-quartile range for the WHOQOL-BREF. In general, the mean is a measure of central tendency, while the inter-quartile range describes the spread of the middle half of scores. A mean can be useful for comparing average scores, but it may not describe the distribution fully. The registry's choice to report an inter-quartile range alongside a mean gives information about spread, but no observed summary values are reported in the registry record.

A confidence interval, in general, expresses uncertainty around an estimated quantity under a model and its assumptions. A narrow interval indicates greater statistical precision than a wide interval, but precision is distinct from freedom from bias. For a nonrandomized comparison, a precise estimate could still be affected by confounding or selection. No confidence interval for the primary outcome is reported in the registry record.

Why does clustering matter?

The intervention was delivered to nursing homes, while outcomes concern people associated with those homes. In general, participants within the same home may share care practices and local conditions, so their outcomes may be more alike than outcomes from people in different homes. This dependence is called clustering. Analyses that treat every participant as fully independent can understate uncertainty when clustering is present. A suitable analysis may account for the home as a cluster, but the registry record does not report whether clustering was addressed or what statistical model was used.

Generalized to count outcomes, over-dispersion occurs when observed variability exceeds what a simple model assumes. A model that fails to account for over-dispersion can produce misleading uncertainty estimates. This concept is not a reported feature of the primary continuous quality-of-life outcome, and the registry record does not provide enough analysis detail to determine whether it arose in any secondary analysis.

How should planning illustrations be interpreted?

The registry's sample-size assumptions describe a planning illustration based on an observed effect, using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The assumed effect to detect is stated as a fraction of the observed effect. These are planning assumptions, not evidence that a planned sample size was achieved or that the study had a particular probability of detecting an effect. Such planning values should never be described as observed power. The registry record does not report the trial's own sample-size justification, and the underlying observed outcome effect is not reported in the registry record.

For time-to-event planning, the record says totals refer to events; for other planning illustrations, values refer to participants per arm. These conventions clarify what a planning quantity represents, but do not supply an outcome analysis. Since no primary endpoint estimate is reported, the size and precision of any intervention-control difference cannot be recovered from the record.

Finally, a listed outcome measure does not guarantee that it was analyzed or that results were reported. The registry lists 8 outcome measures and 0 analyses. The record does not identify a statistical test for the primary endpoint, describe handling of repeated measurements, or state whether analyses were exploratory or confirmatory. Statistical interpretation must therefore remain limited to the design and measurement information that is documented.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 8 outcome measures, including the WHOQOL-BREF primary outcome and the Person-centred Care Assessment Tool (P-CAT) staff version and Person-Centred Climate Questionnaire (PCQ-S) as secondary outcomes. It does not report a multiplicity strategy, such as how interpretation would account for testing several outcomes or subscales. The record lists 0 analyses, so it does not report which outcomes were tested, which comparisons were made, or whether any adjustment for multiple testing was used.

In general, when many outcomes or comparisons are examined, the chance of observing a nominally unusual result somewhere among them can increase. A prespecified hierarchy or an adjustment can help control the interpretation of multiple tests. The registry record does not report whether such procedures were planned or applied. This uncertainty does not establish that a particular result is false; it means the testing context cannot be assessed from the record.

Missing data and analysis populations are also not reported in the registry record. It does not give the number of older people with WHOQOL-BREF data at 9 months, describe reasons for missing measurements, or state how missing values were handled. It does not specify whether the primary analysis included all eligible participants, only those with observed scores, or another defined population. No imputation method, sensitivity analysis, or follow-up completion information is reported.

In general, missing outcome data can affect both precision and validity, particularly if the likelihood of missingness is related to health, participation, or the outcome itself. Different approaches rely on different assumptions. Without details of missingness and analysis populations, readers cannot assess whether the reported target population matches those represented in an analysis; in this record, no primary analysis results are reported at all.

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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: death (47.6%).

The registry records serious and other adverse events separately for older people, next of kin, staff, managers, and staff focus groups. For older people, the intervention group entry lists 42 at risk, 20 with serious adverse events, 42 at risk for other adverse events, 0 with other adverse events, and 20 deaths. The control group entry lists 40 at risk, 19 with serious adverse events, 40 at risk for other adverse events, 0 with other adverse events, and 19 deaths. These are registry entries, not estimates of an intervention effect.

For next of kin, the intervention entry lists 152 at risk and 62 with other adverse events; the control entry lists 156 at risk and 51 with other adverse events. Neither entry lists serious adverse events or deaths. Among staff, the intervention entry lists 363 at risk, 1 with a serious adverse event, 195 with other adverse events, and 1 death; the control entry lists 330 at risk, 0 serious adverse events, 132 with other adverse events, and 0 deaths. Manager and staff focus-group entries list no adverse events or deaths.

The registry identifies Death as the top serious event term at 47.6%. Its denominator and derivation are not reported in the registry record. The recorded deaths should not be attributed to the intervention on the basis of these entries alone. The record does not report event-level timing, causality assessments, or a comparative safety analysis. Group-specific denominators and roles differ, so counts should be read in their own group context rather than combined.

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

The registry record establishes that a nursing-home educational intervention was studied using a nonrandomized, sequential design. It identifies quality of life at 9 months as the primary outcome and describes the WHOQOL-BREF scale, including its direction and score ranges. These details clarify what the study intended to measure and the context in which its results would need interpretation.

The record does not report the outcome estimates needed to determine whether quality-of-life scores differed between intervention and control groups. There are no reported group means, inter-quartile ranges, effect estimates, confidence intervals, or p-values for the primary outcome. Therefore, the record cannot support a conclusion about the direction, magnitude, or precision of any difference. It also cannot establish that the intervention improved or worsened quality of life.

Even if an outcome estimate were available, the nonrandomized allocation would require careful interpretation. In general, observed differences could reflect baseline differences between nursing homes or participants, changes over time, or the intervention. The registry does not report whether the analysis adjusted for such factors. The absence of reported methods and results prevents evaluation of how these alternative explanations were addressed.

The listed enrollment of 1151 should not be mistaken for the number contributing primary-outcome data. The registry does not report endpoint-specific participation, missingness, or analysis populations. Likewise, safety counts describe recorded events in different participant groups, but do not on their own show that the program caused those events or establish a comparative safety profile.

The project also had implementation aims involving staff, leadership, barriers, and facilitators. Those questions require evidence about implementation processes and context; they cannot be answered by the primary quality-of-life endpoint alone. The registry record describes those aims but does not report their analytical findings. The appropriate conclusion is limited: the record documents the study design and planned outcome framework, while its outcome and implementation results are not reported in the registry record.

13. Why This Trial Matters Statistically

This study is statistically notable because it evaluates an educational program delivered in nursing homes, while its primary outcome concerns individual quality of life. That structure raises a general design issue: an intervention applied at the home level can create correlated outcomes among people receiving care in the same setting. Analyses may need to account for that clustering to represent uncertainty appropriately. The registry record does not say whether or how it was addressed.

Nonrandomized allocation also makes the choice and transparency of analysis important. Intervention and control homes may differ in ways that influence both implementation and quality of life. A clear account of baseline characteristics, adjustment variables, and sensitivity analyses would help readers distinguish observed association from a more credible estimate of intervention effect. Such analytical details and results are not reported in the registry record.

The project asks questions beyond a single patient outcome. It includes next of kin, staff, managers, and staff focus groups, alongside measures of person-centred care and climate. This breadth reflects that implementation may involve both care recipients and the people and organizations around them. Statistically, however, outcomes from these groups represent different constructs and populations; they should not be collapsed into a single summary without a defined rationale.

The registry defines the primary outcome and lists outcome measures, but lists 0 analyses. This difference between an outcome framework and reported evidence is important for readers: a registered measure shows what was intended to be assessed, not what the data demonstrated. Reporting estimates, uncertainty, missingness, and analysis methods would be necessary to evaluate the study's findings. Without them, the record contributes design context but not a quantitative estimate of the intervention's effect.

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 central limitation for interpreting this trial record is the absence of reported outcome analyses. Although the primary outcome and timeframe are described, observed scores and comparative estimates are not reported in the registry record. The record lists 0 analyses, so it is not possible to assess statistical significance, effect magnitude, or precision.

The nonrandomized allocation limits causal interpretation. The registry record does not report the allocation process, baseline comparability, or whether the primary analysis adjusted for differences between homes or participants. The sequential model is named, but details of sequence timing and handling of time-related changes are not reported. No masking was used, which is a recorded design feature; the record does not explain how outcome measurement or implementation behavior may have been affected.

The intervention was delivered in nursing homes, but the registry does not report whether analysis accounted for clustering by home. It also does not report the endpoint-specific number of participants, missing-data patterns, follow-up completeness, or analysis populations. These omissions prevent assessment of how representative the measured outcomes were of the intended population.

The primary outcome includes a total score and several subscales, and the registry lists multiple outcome measures. A statistical plan for multiplicity, outcome hierarchy, or subgroup analysis is not reported. The record also does not report a statistical test or model for the endpoint, model diagnostics, or sensitivity analyses.

The adverse-event entries are group-specific, but the registry does not report event timing, relatedness assessments, or a comparative safety analysis. The listed enrollment cannot substitute for endpoint-specific data or clarify the population contributing to each outcome. Taken together, the record supports description of the study framework, but leaves important design, analysis, and outcome details unavailable for appraisal.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Implementation of Knowledge-Based Palliative Care — NCT02708498.
• Registry record last updated 2020-10-28; status last verified 2020-10.
• Results first posted 2020-09-02.
• 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

NCT02708498, “Implementation of Knowledge-Based Palliative Care,” is recorded as completed. The study was sponsored by Lund University and used nonrandomized allocation, a sequential model, and no masking. The registry lists enrollment of 1151 and identifies an educational behavioral intervention delivered in nursing homes in Kronoberg and Skåne, with control homes receiving no intervention.

The primary outcome was WHOQOL-BREF quality of life at 9 months for older people living in nursing homes. The listed inclusion criteria were age 65 years or older, Swedish language, no dementia, and sufficient energy for a structured interview of up to one hour. Secondary outcomes included staff versions of P-CAT and PCQ-S.

The registry lists 8 outcome measures and 0 analyses. It does not report primary-outcome scores, comparative estimates, confidence intervals, p-values, missing-data handling, or analysis populations. Adverse-event counts are recorded for older people, next of kin, staff, managers, and staff focus groups, but the record does not report a comparative safety analysis or establish event causality.