This study used the Healthcare Cost and Utilization Project (HCUP) National Inpatient Sample (NIS) covering 2007–2011 as the data source. From that database, investigators identified 802,080 primary admissions coded for lower gastrointestinal (GI) bleeding using ICD-9 diagnosis codes. Of these admissions, 283,925 patients had documented cardiovascular comorbidities. The NIS is a third-party, nationally representative inpatient dataset maintained by the Agency for Healthcare Research and Quality (AHRQ); access requires completion of HCUP training and a data use agreement.
The stated aim was to evaluate whether patients with cardiovascular comorbidities, many of whom may be on antiplatelet or anticoagulation therapy, are more likely to be diagnosed with new benign colorectal neoplasms or malignant colorectal neoplasms during an admission for lower GI bleed than patients without cardiovascular comorbidities. The analysis focused on the incidence of newly recorded colorectal neoplasms in the context of lower GI bleeding.
Investigators selected all primary lower GI bleed admissions from the NIS (2007–2011) using ICD-9 codes. The cohort was stratified into two groups: those with cardiovascular comorbidities and those without. The primary outcome was a new diagnosis of either benign or malignant colorectal neoplasm recorded during that hospitalization. The abstract does not list the specific ICD-9 codes used to define colorectal neoplasms, nor does it provide a detailed list of which cardiovascular conditions were included in the comorbidity definition. Medication exposure (for example, antiplatelet or anticoagulant use) was discussed as context but specific medication data or adjustments were not reported in the abstract.
Weighted analyses accounting for the NIS complex survey design were performed using SAS 9.4. Temporal trends across 2007–2011 were assessed with Cochran–Armitage trend tests and linear regression. Associations between cardiovascular comorbidity status and colorectal cancer stage were evaluated by multinomial logistic regression. Statistical significance was defined as p ≤ 0.05. The abstract does not provide model covariates, effect estimates beyond the odds percentage changes cited, confidence intervals, or p-values for specific comparisons other than describing the percentage differences and noting significance for the 2010 benign neoplasm finding.
Across the 2007–2011 period, the authors report that patients who presented with a lower GI bleed and did not have cardiovascular comorbidities had higher odds of being diagnosed with malignant colorectal neoplasms compared with those who had cardiovascular comorbidities. The reported increase in odds for malignant neoplasms among patients without cardiovascular comorbidities ranged from 49.9% to 65.3% over the study years. The abstract presents these percentage increases as the primary contrast for malignant disease between groups; detailed year-by-year estimates, confidence intervals, absolute rates, and multivariable-adjusted effect sizes are not provided in the abstract.
The authors highlight a notable year-specific difference in 2010: patients with cardiovascular comorbidities had 10.5% higher odds of being diagnosed with benign colorectal neoplasms compared with patients without cardiovascular comorbidities, and this difference was reported as statistically significant. No further year-by-year breakdowns, absolute prevalences of benign neoplasms, or adjustments for potential confounders are described in the abstract.
Based on the NIS analysis of 2007–2011 admissions for lower GI bleeding, the authors conclude that patients without cardiovascular comorbidities were significantly more likely to be diagnosed with malignant colorectal neoplasms than patients with cardiovascular comorbidities. They also report a specific 2010 finding where benign neoplasms were more likely to be diagnosed in patients with cardiovascular disease. The abstract frames possible relevance of antiplatelet or anticoagulant therapy among cardiovascular patients as background but does not present medication-specific results.
The authors declared no competing interests and stated that ethical guidelines were followed; the abstract indicates IRB/oversight details were recorded as not applicable. They confirm appropriate reporting standards and that any necessary consent processes were handled. The underlying data come from the HCUP NIS, which is accessible to qualified users through the HCUP Central Distributor under a Data Use Agreement; the authors state they did not receive special access. The abstract does not report detailed limitations, adjustment variables, diagnostic code lists, or full model outputs; because this is a medRxiv preprint, the work had not been peer-reviewed at the time of posting and should not be used to guide clinical practice.
Note: Details such as specific ICD-9 code lists, adjustment covariates, confidence intervals, absolute rates, and the full set of statistical outputs were not reported in the abstract and therefore are not included here.