On logistic regression with right censored data, with or without competing risks, and its use for estimating treatment effects

Propensity weighting plus adjustment in proportional hazards model is not doubly robust

Recently, it has become common for applied works to combine commonly used survival analysis modeling methods, such as the multivariable Cox model and propensity score weighting, with the intention of forming a doubly robust estimator of an exposure effect hazard ratio that is unbiased in large samples when either the Cox model or the propensity score model is correctly specified. This combination does not, in general, produce a doubly robust estimator, even after regression standardization, when there is truly a causal effect. We demonstrate via simulation this lack of double robustness for the semiparametric Cox model, the Weibull proportional hazards model, and a simple proportional hazards flexible parametric model, with both the latter models fit via maximum likelihood. We provide a novel proof that the combination of propensity score weighting and a proportional hazards survival model, fit either via full or partial likelihood, is consistent under the null of no causal effect of the exposure on the outcome under particular censoring mechanisms if either the propensity score or the outcome model is correctly specified and contains all confounders. Given our results suggesting that double robustness only exists under the null, we outline 2 simple alternative estimators that are doubly robust for the survival difference at a given time point (in the above sense), provided the censoring mechanism can be correctly modeled, and one doubly robust method of estimation for the full survival curve. We provide R code to use these estimators for estimation and inference in the supporting information.

Cannabis for chronic pain: cardiovascular safety in a nationwide Danish study

BACKGROUND AND AIMS

A rising number of countries allow physicians to treat chronic pain with medical cannabis. However, recreational cannabis use has been linked with cardiovascular side effects, necessitating investigations concerning the safety of prescribed medical cannabis.

METHODS

Using nationwide Danish registers, patients with chronic pain initiating first-time treatment with medical cannabis during 2018-21 were identified and matched 1:5 to corresponding control patients on age, sex, chronic pain diagnosis, and concomitant use of other pain medication. The absolute risks of first-time arrhythmia (atrial fibrillation/flutter, conduction disorders, paroxysmal tachycardias, and ventricular arrhythmias) and acute coronary syndrome were reported comparing medical cannabis use with no use.

RESULTS

Among 1.88 million patients with chronic pain (46% musculoskeletal, 11% cancer, 13% neurological, and 30% unspecified pain), 5391 patients claimed a prescription of medical cannabis [63.2% women, median age: 59 (inter-quartile range 48-70) years] and were compared with 26 941 control patients of equal sex- and age composition. Arrhythmia was observed in 42 and 107 individuals, respectively, within 180 days. Medical cannabis use was associated with an elevated risk of new-onset arrhythmia {180-day absolute risk: 0.8% [95% confidence interval (CI) 0.6%-1.1%]} compared with no use [180-day absolute risk: 0.4% (95% CI 0.3%-0.5%)]: a risk ratio of 2.07 (95% CI 1.34-2.80) and a 1-year risk ratio of 1.36 (95% CI 1.00-1.73). No significant association was found for acute coronary syndrome [180-day risk ratio: 1.20 (95% CI 0.35-2.04)].

CONCLUSIONS

In patients with chronic pain, the use of prescribed medical cannabis was associated with an elevated risk of new-onset arrhythmia compared with no use-most pronounced in the 180 days following the initiation of treatment.

Treated periodontitis and recurrent events after first-time myocardial infarction: A Danish nationwide cohort study

AIM

To investigate the association between previous periodontal treatment and recurrent events after first-time myocardial infarction (MI).

MATERIALS AND METHODS

From the Danish nationwide registries, patients with first-time MI between 2000 and 2015 were divided into three groups according to oral health care within 1 year prior to first-time MI. A multiple logistic regression model provided adjusted odds ratios (ORs) with 95% confidence intervals (CIs) to assess the 3-year risk of major adverse cardiovascular events (MACE).

RESULTS

A total of 103,949 patients were included. Patients with treated periodontitis (PD) prior to first-time MI had an adjusted 3-year risk of MACE similar to patients presumed periodontally healthy (OR 0.97 [95% CI 0.92-1.03]). Patients with no prior dental visits were significantly older, had more comorbidities and showed significantly increased adjusted 3-year risks of MACE (OR 1.47 [95% CI 1.42-1.52]), cardiovascular death (OR 1.71 [95% CI 1.64-1.78]) and heart failure (OR 1.13 [95% CI 1.07-1.20]) compared with patients presumed periodontally healthy.

CONCLUSIONS

Patients with treated PD 1 year prior to first-time MI had a similar risk of recurrent cardiovascular events as patients presumed periodontally healthy. No dental visit prior to first-time MI was an independent risk factor for recurrent events.

The effect of discontinuing beta-blockers after different treatment durations following acute myocardial infarction in optimally treated, stable patients without heart failure: a Danish, nationwide cohort study

AIMS

We studied the effect of discontinuing beta-blockers following myocardial infarction in comparison to continuous beta-blocker use in optimally treated, stable patients without heart failure.

METHODS AND RESULTS

Using nationwide registers, we identified first-time myocardial infarction patients treated with beta-blockers following percutaneous coronary intervention or coronary angiography. The analysis was based on landmarks selected as 1, 2, 3, 4, and 5 years after the first redeemed beta-blocker prescription date. The outcomes included all-cause death, cardiovascular death, recurrent myocardial infarction, and a composite outcome of cardiovascular events and procedures. We used logistic regression and reported standardized absolute 5-year risks and risk differences at each landmark year. Among 21 220 first-time myocardial infarction patients, beta-blocker discontinuation was not associated with an increased risk of all-cause death, cardiovascular death, or recurrent myocardial infarction compared with patients continuing beta-blockers (landmark year 5; absolute risk difference [95% confidence interval]), correspondingly; -4.19% [-8.95%; 0.57%], -1.18% [-4.11%; 1.75%], and -0.37% [-4.56%; 3.82%]). Further, beta-blocker discontinuation within 2 years after myocardial infarction was associated with an increased risk of the composite outcome (landmark year 2; absolute risk [95% confidence interval] 19.87% [17.29%; 22.46%]) compared with continued beta-blocker use (landmark year 2; absolute risk [95% confidence interval] 17.10% [16.34%; 17.87%]), which yielded an absolute risk difference [95% confidence interval] at -2.8% [-5.4%; -0.1%], however, there was no risk difference associated with discontinuation hereafter.

CONCLUSION

Discontinuation of beta-blockers 1 year or later after a myocardial infarction without heart failure was not associated with increased serious adverse events.