Clinical Characteristics and Disease Burden of Diabetic Patients Based on Tianjin Regional Database: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Clinical Characteristics and Disease Burden of Diabetic Patients Based on Tianjin Regional Database (NCT04903496), a registered phase not stated study. The primary outcome is Number of Participants With Established Cardiovascular Disease, or Chronic Kidney Disease, or High Cardiovascular Risk.

2. Clinical Question

3. Trial Design

This completed study is a retrospective analysis of electronic healthcare records in the Tianjin regional database. The registry describes a population of Chinese diabetic and non-diabetic inpatients and outpatients, organized according to whether diagnoses of CVD, HF, CKD, or high cardiovascular risk were recorded. The study start date is 2021-09-06, and the primary completion date is 2022-06-24. The registry lists enrollment of 1233162.

The listed arms separate patients by diabetes status and recorded risk-factor status. For the primary outcome, the diabetic categories are combined and the non-diabetic categories are combined. This makes the primary comparison a descriptive comparison between groups identified from records. Allocation is not stated, and the registry specifically does not describe the arms as a like-for-like control comparison. No treatment assignment or randomized allocation is reported.

The primary outcome covers records from 2015 up to the given year, with the stated timeframe extending through 2019. Secondary outcomes include age, sex counts, and insurance-payment counts for inpatients and all inpatients and outpatients at their last visit in specified years. The registry lists 25 outcome measures and 0 analyses. That does not establish whether analyses were performed outside the registry record; their details are not reported in the registry record. The record does not give a participant flow, detailed eligibility criteria, or a protocol description beyond the stated objectives and outcome definitions.

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: Only participants that met all inclusion and none of the exclusion criteria were included. Participants with at least 1 diagnosis record and baseline information (birth date and sex) were identified in the Tianjin regional medical database between 01/01/2015 to 31/12/2019 and allocated in 4 (A,B,C, and D) arms. The 4 arms were mutually exclusive in each year (2015, 2017, and 2019) but not across years.

Study Period 2015

Study Period 2017

Study Period 2019

Analysis population for the primary outcome, as defined in the registry: In- and outpatients in 2015, 2017, and 2019 with and without diabetes and with and without diagnosis of risk factors CVD, HF, CKD, and high CV risk. For this outcome, groups were combined as per protocol to 'Diabetic Patients \[Arm A and B\]' and 'Non-Diabetic Patients \[Arm C and D\]'. Arms were not mutually exclusive. Only patients with available data were included in this analysis.

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: Number of participants with at least one recorded International Classification of Disease (ICD) code diagnosis in the Tianjin regional database, Chinese nationality, and baseline information (birth date and sex) in 2015, 2017, and 2019.

6. Primary Endpoint

The registry lists the primary outcome as Number of Participants With Established Cardiovascular Disease, or Chronic Kidney Disease, or High Cardiovascular Risk, measured over up to 5 years (2015 up to 2019).

The primary endpoint is titled as a count of participants with established CVD, CKD, or high cardiovascular risk. Its description specifies a calculation based on patients with labels for the diseases of interest from 2015 through a given year, divided by diabetic or non-diabetic patients from 2015 through that year and expressed as a percentage. Thus, the endpoint description contains both a count-based title and a period-based proportion calculation. The actual endpoint values and denominators are not reported in the registry record.

The endpoint is intended to describe the burden of recorded diagnoses in the database and to compare diabetic with non-diabetic patients within and across the specified years. Disease labels were identified using International Classification of Diseases codes. The record notes that patients could have multiple diagnoses at the same time, so CVD, HF, CKD, and high cardiovascular risk should not be assumed to form mutually exclusive categories.

This measure is useful for describing how often diagnoses appear in a defined healthcare database population. It does not by itself establish that diabetes caused a diagnosis, that the database represents all people in the region, or that one group’s underlying risk differs because of diabetes. The registry does not report the detailed coding rules, group denominators, or uncertainty estimates needed to assess the endpoint’s precision.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a retrospective electronic-record analysis. In this design, researchers use information already recorded during healthcare rather than assigning an intervention for the purpose of the study. The listed groups are formed from diagnoses and diabetes status. Because allocation is not stated and there is no randomized treatment assignment described, the groups are not equivalent to randomized trial arms. Differences between them could reflect differences in health status, care-seeking, diagnosis, or other factors; the registry record does not report methods that address these possibilities.

How is the primary measure defined?

The endpoint title refers to a number of participants, while its description gives a percentage calculation: patients with diagnoses of interest recorded from 2015 through a given year are related to the diabetic or non-diabetic patient denominator from 2015 through that year. This is a period-based summary, not necessarily a measure of newly occurring disease during a particular year. A person’s recorded diagnosis history, the period covered, and the way patients enter the denominator can affect interpretation. The registry does not report enough detail to determine the precise handling of repeated visits, changing diagnoses, or patients appearing in more than one year.

International Classification of Diseases codes were used to identify diagnoses of interest. Coding can make large records suitable for consistent classification, but the results depend on which codes are included and how accurately diagnoses are recorded. The specific code lists and validation procedures are not reported in the registry record. The registry also notes that patients could have several diagnoses simultaneously. Accordingly, category counts should not be presumed to add up to a total, and overlapping conditions may be represented in more than one category.

What can a proportion tell the reader?

In general, a proportion describes the share of a defined denominator meeting a stated condition. Its interpretation depends on a clearly specified numerator, denominator, and observation period. A confidence interval, in general, expresses statistical uncertainty around an estimate under the assumptions of the method used; it does not capture every source of bias or data-quality problem. No confidence intervals or other uncertainty measures for the primary outcome are reported in the registry record, so the precision of the estimates cannot be assessed from this record.

What about comparisons and variation in the data?

In general, an unadjusted comparison summarizes a difference between observed groups without controlling for other characteristics. An adjusted analysis attempts to account for selected factors, but its validity depends on which factors are measured and how the model is specified. The registry record does not report an adjusted model, a test for group differences, or an approach for repeated observations. It therefore does not support a conclusion about whether any apparent group contrast is statistically distinguishable from sampling variation or reflects other differences between groups.

In general, over-dispersion occurs when the observed variation in data is greater than a statistical model expects. It can make standard errors too small if a model assumes less variation than the data contain. The registry record does not report a model or dispersion assessment, and there is no basis to claim that over-dispersion was present or handled in this analysis. The registry’s endpoint is also a database-based descriptive measure; model-based methods would require details about the unit of analysis, repeated records, and denominator construction that are not reported.

How should planning illustrations be read?

Any sample-size illustrations are planning scenarios, not observed power or evidence about the study’s design. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect set as a stated fraction of an observed effect. The registry record does not report that these assumptions were used to design the study, nor does it provide a sample-size justification. Since this analysis is not a randomized comparison and its primary results are not reported, such illustrations should not be taken as a formal assessment of the study’s ability to detect a clinical effect.

10. Multiplicity, Missing Data and Analysis Populations

The study lists 25 outcome measures, including the primary outcome and secondary measures concerning age, sex, and insurance payment. When many outcomes or comparisons are examined, the chance of finding an apparently notable result by chance can increase. In general, multiplicity methods may control that risk by adjusting how results are evaluated or by defining a limited set of confirmatory hypotheses in advance. The registry record does not report a multiplicity strategy, a hierarchy among secondary outcomes, or which comparisons were considered confirmatory.

Missing-data procedures are not reported in the registry record. The record does not explain how missing diagnoses, absent demographic information, missing insurance records, or incomplete visits were handled. It also does not report how patients with partial records were treated in the endpoint denominator. These details matter because incomplete recording can affect both the number classified with a diagnosis and the population used to calculate a proportion.

The primary endpoint definition identifies combined diabetic and non-diabetic groups, but analysis-population rules beyond that are not reported in the registry record. There is no reported accounting of exclusions, eligibility decisions, record linkage, or the numbers contributing to each analysis. The registry lists 0 analyses; this does not establish that no analyses were conducted, but it means their specifications cannot be evaluated from the record. There is also no reported statistical analysis plan, interim-analysis approach, or handling of repeated observations.

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

The registry record establishes what the study set out to describe: recorded CVD, CKD, and/or high cardiovascular risk among diabetic and non-diabetic patients in a regional healthcare database. It also gives the broad study period, group definitions, and the use of diagnosis codes. These features make the intended descriptive comparison understandable.

The record does not report the primary outcome estimates or their denominators. It therefore cannot show how common the diagnoses were in either group, how the groups differed, or how stable the estimates were. It also does not provide statistical uncertainty or adjusted analyses. No claim about statistical significance, the size of an association, or the direction of a difference can be supported by the record alone.

Even if outcome values were available, an observational comparison of recorded diagnoses would not establish that diabetes caused those diagnoses. The database reflects documented healthcare encounters, so its records may depend on who received care, which assessments were performed, and how diagnoses were coded. The registry does not report enough detail to quantify these influences or establish how representative the database is of the broader population.

The count entries for deaths and other adverse-event categories likewise do not establish rates or comparative risks without denominators and fuller outcome definitions. The study is marked completed, but completion status is not a substitute for numerical results or a statistical analysis plan. The appropriate conclusion is limited: the registry describes a retrospective database study and its intended endpoints, while the numerical evidence needed to judge the primary findings is not reported in the registry record.

12. Why This Trial Matters Statistically

This study is statistically notable for the distinction between a registry-defined group and a randomized comparison. Patients are categorized by recorded diabetes and risk-factor diagnoses, and the primary outcome combines categories by diabetes status. This can support descriptive epidemiology, but it does not ensure that the compared groups are similar in other relevant characteristics. The absence of reported adjustment methods limits interpretation of any group contrast.

The endpoint also illustrates why the measure’s title and calculation both matter. A title framed as a participant count is accompanied by a percentage calculation covering diagnoses recorded from 2015 through a given year. Understanding whether a result is a count, a period-based proportion, or a year-specific rate depends on its numerator, denominator, and time definition. Here, the registry gives a broad definition, but endpoint values and denominators are not reported.

Finally, the record lists a large enrollment figure and many outcome measures, but those details do not by themselves demonstrate precision, representativeness, or a statistically persuasive result. Precision depends on the number of relevant observations in each analysis and on the outcome’s variability; representativeness depends on how the database population relates to the population of interest. The lack of reported results, uncertainty estimates, and missing-data and multiplicity methods makes those questions difficult to assess. This is a useful example of why study design, outcome definition, and transparent reporting must be considered alongside headline enrollment.

13. 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 interpretation is that the registry record does not report the primary outcome values, endpoint-specific denominators, or uncertainty estimates. As a result, neither the magnitude nor the precision of the intended diabetic versus non-diabetic comparison can be assessed. Although the record lists enrollment of 1233162, it does not explain how enrollment maps to the observations contributing to each outcome or year.

The study relies on retrospective healthcare records and diagnosis codes. The registry record does not report the code lists, coding validation, or procedures for handling inconsistent or repeated records. Because a patient could have multiple diagnoses simultaneously, the categories may overlap. The record does not explain how repeated visits or patients present in multiple years affect the calculations.

Eligibility criteria, missing-data handling, and the construction of analysis populations are not reported in sufficient detail to assess selection or information bias. The registry does not report adjustment for potential differences between groups, a multiplicity strategy, or a statistical analysis plan. Nor does it give enough information to judge how representative the regional database is of the broader Chinese population. The recorded death counts lack the denominators and additional context required for rate comparisons. These limitations concern what can be concluded from the registry record; they do not establish that the underlying analysis lacked further methods or results.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Clinical Characteristics and Disease Burden of Diabetic Patients Based on Tianjin Regional Database — NCT04903496.
• Registry record last updated 2024-11-12; status last verified 2024-09.
• Results first posted 2024-11-12.
• 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.

17. Record Summary

NCT04903496 is a completed retrospective analysis of electronic healthcare records from the Tianjin regional database. The study concerns Chinese inpatients and outpatients with and without diabetes, classified by recorded CVD, HF, CKD, or high cardiovascular risk. The registry lists enrollment of 1233162, a start date of 2021-09-06, and primary completion on 2022-06-24.

The primary outcome concerns participants with established CVD, CKD, and/or high cardiovascular risk, with groups combined by diabetes status and diagnoses identified using International Classification of Diseases codes. Secondary outcomes include age, sex, and insurance-payment measures. The registry lists 25 outcome measures and 0 analyses. Primary numerical results, denominators, uncertainty estimates, detailed analysis methods, and missing-data procedures are not reported in the registry record. The study is observational; its groupings should not be read as randomized treatment comparisons.