Nabagiez 2013.

Study characteristics
Methods	RT Study duration: 13 months
Participants	701 participants Setting: home Diagnostic criteria: all participants discharged to home following coronary artery bypass graft procedure and/or valve repair or replacement and/or aneurysm repair, or other cardiac procedure Age (years) (mean ± SD): intervention group: 62.8 (10.6), control group: 63.2 (10.9) Sex female n (%): intervention group: 73 (21.5%), control group: 88 (24.4%) Country: USA Comorbidity n (%): diabetes mellitus: intervention 123 (34.0), control 111 (32.6); hypertension: intervention 268 (74.2), control 283 (83.2); dyslipidaemia: intervention 263 (72.8), control 274 (80.5); dialysis: intervention 8 (2.2), control 7 (2.0); cerebrovascular accident: intervention 15 (4.1), control 9 (2.6); COPD: intervention 44 (12.1), control 30 (8.8); peripheral vascular disease: intervention 29 (8.0); control 25 (7.3); previous myocardial infarction: intervention 146 (40.4), control 144 (42.3); CHF: intervention 51 (14.1), control 48 (14.1); arrhythmia: intervention 37 (10.2), control 39 (11.4) Sociodemographics: not stated Ethnicity: intervention group: 289 (84.4%) white, 53 (15.5%) non‐white; control group: 318 (88%) white, 43 (11.9%) non‐white Date of study: August 2009‐September 2011
Interventions	1 intervention group 340 participants, control 361 participants Hospital‐employed, cardiothoracic physician assistants conducted home visits on post‐discharge days 2 and 5, with occasional variation due to participant availability and Sundays, on which no house calls were made. The same hospital‐based physician assistants responsible for perioperative and intraoperative care were assigned to make house calls. During a house call, the physician assistant performed a focused physical exam and reviewed the participant’s medications. Adjustments were made to the participant’s medications, and new medications were prescribed as necessary. The surgical wounds were examined carefully and all participant concerns were addressed. Prescriptions were written for antibiotics, blood work, or imaging studies when indicated. Arrangements were made if the participant needed to be evaluated as an inpatient. All findings were documented on the visit form. Both groups were seen in the office on post‐discharge weeks 2 and 4. The control group was seen at home by standard visiting nurses without any specialty training or expertise in caring for people with cardiac surgery
Outcomes	Hospital admissions/number of people admitted to hospital
Notes	Not sure of randomisation and these were hospital‐based physician assistants working in homes.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Unclear from the document
Allocation concealment (selection bias)	Unclear risk	Unclear from the document
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Unclear from the document
Incomplete outcome data (attrition bias) All outcomes	High risk	19% of the participants in the intervention group refused to participate or failed to respond to requests to participate.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	This is less important for our purpose as we are looking at objective outcomes (hospitalisations, ED visits, and mortality).
Selective reporting (reporting bias)	Low risk	No evidence of this
Protection against contamination bias	Low risk	No contamination
Other bias	Unclear risk	Unclear. There is no evidence of other bias