Atrial fibrillation (AF) is one of the most common arrhythmias with significant morbidity and mortality. Similar to previous studies, we defined new-onset secondary AF (NOSAF) as initial diagnosis of AF (without preexisting AF) that occurred during medical management of an unrelated noncardiac condition. A recent study showed increased prevalence of device detected AF around the time of noncardiac surgery or medical illness requiring hospitalization with half the patients having similar episodes in the past. Estimates of NOSAF occurrence in the critically ill population range up to 44%. One school of thought is that NOSAF occurs as an isolated event in the setting of acute illness and does not portend a poor prognosis in the long-term. Importantly, the general opinion is that NOSAF does not merit the same management strategies as a history of paroxysmal or persistent AF. This is especially relevant in noncardiac critical illness. With an emerging interest in NOSAF, we sought to analyze the outcomes in studies that evaluated NOSAF in noncardiac critically ill patients.

We included studies assessing noncardiac critically ill adult patients who developed NOSAF and had a comparator group with either no previous AF or previous AF presenting with or without AF in the index hospitalization. We excluded patients admitted for cardiovascular or thoracic surgery as they are known risk factors for the development of AF. The outcome data had to include either in-hospital mortality, hospital length of stay (LOS), intensive care unit (ICU) LOS, long-term mortality (≥1 year), or stroke. Articles had to be published in English and have full text available.

To identify eligible studies for inclusion in the systematic review and meta-analysis, 3 independent reviewers (KS, LS, and MK) systematically searched Cochrane CENTRAL, PubMed, United Kingdom Clinical Trials Gateway, and the WHO International Clinical Trials Registry for relevant studies. Search terms were “secondary,” “new-onset,” or “recent onset;” “atrial fibrillation,” “atrial flutter,” or “arrhythmia;” and “sepsis,” “critically ill,” “pneumonia,” or “acute disease”. In addition, we reviewed citations from the eligible studies. This review followed the MOOSE guidelines for meta-analysis reporting. Available studies were screened based on the title and abstract. Full text articles were obtained for studies that could be eligible. The eligibility of the articles was determined by the same reviewers by going through each article (n = 59). We collected data on author, year, study design, outcomes measured, follow-up rate. We collected absolute event counts. When studies reported matched and unmatched data, we collected both sets of data but used matched data for the analysis. Differences in study inclusion were resolved after discussion and consensus between the 3 reviewers or by a 4th reviewer, (NM), in case of disagreement. The study quality was assessed using the Newcastle-Ottawa Quality Assessment Scale for cohort studies, with scores over 7 indicating high quality. Clinical outcomes of interest were in-hospital mortality, hospital LOS, ICU LOS, stroke occurrence, and long-term mortality. Because all the data were extracted from published manuscripts, our study was exempt from institutional review board approval.

Statistical analysis was performed using meta package for R (version 4.0, R Foundation, Vienna, Austria) and RStudio (version 1.2, RStudio, PBC, Boston, Massachusetts). For binary data, Mantel-Haenszel risk ratio (RR) with random-effects model (DerSimonian and Laird method) was used to summarize data between the groups. For continuous data in the studies which were reported as median and interquartile range, we first used the Wan method to estimate the mean and SD. We then calculated the standardized difference in means (SMD) using a random effects model and Cohen’s d to evaluate the data between the groups. Higgins I-squared ( I ² ) statistic was used to assess the heterogeneity. A value of I ² of 0% to 25% represented insignificant heterogeneity, 26% to 50% represented low heterogeneity, 51% to 75% represented moderate heterogeneity, and more than 75% represented high heterogeneity. Sensitivity analysis was performed with leave-one-out method to evaluate influence of 1 study. Publication bias was formally assessed using funnel plots and Egger’s linear regression test of funnel plot asymmetry for outcomes reported in ≥10 studies. A 2-tailed p <0.05 was considered statistically significant.

Of the 2,360 relevant articles identified, 59 studies were adjudicated as relevant and underwent full-text review. Of these, 19 studies ultimately met the inclusion criteria and were included in the meta-analysis. ^, These studies comprised 306,805 participants ( Figure 1 ). A total of 11 studies were retrospective. ^{, , , , , ,} The majority were conducted at a single center and there were 2 large database studies. Detailed study characteristics can be found in Table 1 . All studies were high quality except for the study by Brathwaite and Weissman which had a score of 4 ( Supplementary Table 1 ).

Flow diagram of systematic search of studies.

There were 18,882 patients who developed NOSAF in the 19 included studies. Across all studies, the mean age was 67.79 years. The mean percentage of women was 44%. Congestive heart failure, hypertension, and diabetes mellitus were common, with mean percentages of 22%, 56%, and 31%, respectively. Patient characteristics are summarized in Table 2 .

Table 2

Patient characteristics for the studies at baseline

Study	NOSAF
	n	Age (y)	Female	Hypertension	DM	CHF	CAD	Stroke	Apache
Arrigo et al	212	73	25%	53%	24%	15%	16%	3%	NR
Bedford et al	831	71	40%	NR	NR	NR	NR	NR	NR
Brathwaite et al	47	67	55%	47%	NR	6.3%	26%	NR	8.4
Carrera et al	582	72	43%	68%	32%	13%	26%	12%	60
Chen et al	53	67	51%	NR	34%	25%	28%	NR	27
Cheng et al	1,286	79	47%	35%	26%	31%	17%	NR	NR
Christian et al	18	66	NR	NR	NR	NR	NR	NR	0.49
Fernando et al	1,541	65	45%	34%	28%	15%	NR	NR	NR
Goodman et al	52	69	50%	35%	37%	15%	17%	NR	23
Guenancia et al	29	71	38%	55%	24%	NR	21%	NR	NR
Jacobs et al	213	73	35%	NR	NR	NR	NR	NR	97
Klouwenberg et al	418	66	37%	NR	21%	NR	NR	NR	89
Lewis et al	34	69	53%	79%	44%	47%	24%	41%	NR
Liu et al	240	77	43%	65%	36%	21%	45%	32%	23.7
Makrygiannis et al	20	NR	30%	60%	30%	0	15%	15%	17.9
Moss et al	749	64	45%	66%	32%	19%	31%	12%	NR
Shaver et al	123	70	35%	63%	36%	15%	24%	12%	NR
Walkey et al	2,896	74	44%	46%	28%	11%	5.7%	4.4%	NR
Walkey et al	9,540	81	56%	88%	50%	67%	70%	45%	NR

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Average Exercise Capacity in Men and Women >75 Years of Age Undergoing a Bruce Protocol Exercise Stress Test
2. Layered Plaque Characteristics and Layer Burden in Acute Coronary Syndromes
3. Prognostic Effects of Cardiac Rehabilitation in Patients With Heart Failure (from a Multicenter Prospective Cohort Study)
4. The Sinotubular Junction-to-Aortic Annulus Ratio as a Determinant of Supravalvar Aortic Stenosis Severity*

Stay updated, free articles. Join our Telegram channel

Join