Table 4.

Key findings on primary care post-discharge management of acute coronary syndrome

First Author (Year) Aim or Research question	Key findings on ACS interventions in primary care	Principal conclusions	Study Quality Comments MMAT Score
Cole (2014) [26]	Significant increase (p < 0.01) in frequency of use of all EBM investigated during calendar period 2005–2010 Medication usage at 12 months post-ACS (2010): Aspirin 96 % Clopidogrel 71 % DAPT 68 % βB 70 % ACEi/ARB 80 % Statin 93 %	Guideline-indicated medication use has increased over the 6-year study period, but treatment gap remains	• Data extracted from pre-existing ACS follow-up registry • 89 % follow-up MMAT: 100 %
Fernandez (2006) [38] To investigate risk factor status of post-PCI patients	Risk factor status at 1 year post-PCI: Systolic blood pressure above target 31 % Total cholesterol above target 58 % Smoking 15 % BMI above healthy range 77 % Obesity 34 % Physical activity below target 48 % Depression & anxiety 25 % One third of patients erroneously believed that they had no heart problems	There is inadequate management of identifiable risk factors among post-PCI patients 12–18 months after revascularisation	• 39 % response among eligible participants • Self-reported risk factor status in self-administered questionnaire MMAT: 75 %
Ford (2011) [27] To measure attainment of New Zealand Guideline Group targets & highlight areas of weakness	Risk factor status at 3 years post-ACS (2010): Attainment of target blood pressure 76 % Smokers who quit 52 % BMI in target range 24 % HDL levels above target 74 % LDL levels below target 52 % In 2010, at 3 years post-ACS, % of medications not prescribed by GPs: Aspirin 1 % βB 6 % ACEi/ARB 22 % Statin 3 % GTN spray 27 %	Concern that GPs were using outdated guidelines Mixed achievement of NZGG program—large treatment gaps for BMI, HbA1c & lifestyle Need for further efforts to improve diet & exercise Weight reduction particularly challenging—majority of patients remained overweight/obese	Reports data by ethnicity Survivorship bias (26 patients had died) Interventions implied to be based in primary care MMAT: 75 %
Gallagher (2003) [35] To identify determinants of women’s attendance at CR and adherence to risk factor modification	At 12 weeks post-discharge: • Two-thirds of women referred to CR • Only one third of the total sample attended CR • CABG patients more likely to be referred than MI patients • Lack of employment, age <55 or >70 and stressful personal life event decreased the odds of attending	Good adherence to guidelines on medications, stress modification & smoking Poorer adherence to diet & exercise guidelines	Self-reported outcomes Non-English speakers excluded MMAT: 100 %
Hansen (2011) [43] To investigate experiences of ongoing smoking or smoking cessation post-ACS	In 2006–2008, insights about GP smoking advice to patients post-ACS: • GP advice sometimes resented and sometimes appreciated • GPs more likely to talk to than lecture at patients compared to specialists • Doctor patient rapport is important • Majority of quitters spontaneously quit with no GP advice • Failed quitting attempts lead to hopelessness	Being bombarded with anti-smoking advice during hospitalisation can result in patients “turning off” Anti-smoking advice may have a positive cumulative effect when presented well & at the right time Pharmacotherapy is underutilised GPs could better inform patients about the process of quitting & available supports	• Appropriate subject selection • Low dropout • No comments on how researchers could influence interview responses MMAT: 75 %
Hickey (2004) [30] To determine whether reliable and valid clinical indicators could measure ACS primary and hospital care To determine whether education efforts could improve these clinical indicators	In 2002, insights from a program for hospitals and GPs: • Robust process and outcome clinical indicators can be developed to assess primary and hospital care that are relevant, reliable, valid and high impact • Education program improved 17/40 developed indicators	Suboptimal performance was improved with feedback to GPs. Economical data collection and timely feedback would improve QI process Sustainability of this approach limited by expense and labour	• Listed strategies for minimization of measurement error • Found high accuracy through random sampling of audits MMAT: 100 %
Johnson (2010) [42] To determine whether self-reported receipt of lifestyle advice from a health care provider is lower among outpatient cardiac rehabilitation (OCR) non-attendees and non-referred patients compared to OCR attendees	In 2002–2007, % of patients receiving lifestyle advice from GPs: Advised to increase physical activity 71 % Advised to follow a modified fat diet 55 % Advised to quit smoking out of patients who smoked in last 6 months 88 %	Recommended that referred patients who do not attend CR be identified by their GP and encouraged to participate in home-based CR	• 65 % consented to inclusion • Consenters more likely to be male and undergo CABG • Analysis based on self-report: patient recall 5 months post-discharge • Large sample size: used Hunter New England Heart and Stroke Registry • Potential response bias • May underestimate appropriate advice • Considerable missing data MMAT: 100 %
Looi (2011) [28] To measure adherence to evidence-based ACS medications post-CABG	In 2006–2007, at 3 years post-CABG, % of medication usage by patients: Aspirin 83 % βB 62 % ACEi/ARB 43 % Statins 72 % Major adverse cardiological events (6.2 %/year): 3 UA, 4 NSTEMI, 6 HF, 5 deaths	Secondary prevention medication usage in ACS patients undergoing CABG was disappointingly low at discharge and worse at follow-up	• 86 % response rate • Association between cardiac events and low adherence to cardiac medication was not statistically assessed MMAT: 75 %
Mudge (2001) [33] To measure prescription of lipid-lowering drugs on discharge, and patient adherence at follow up	In 1998–1999, at 6–18 months post-ACS, patient status in lipid management: Did not have lipid measurements 10 % Of patients not prescribed LLD at discharge, patients who did not receive LLD prescription from GP 70 % Of those prescribed LLD on discharge, patients who remained on the treatment at GP follow up 88 %	Identified suboptimal lipid documentation with poor communication across hospital-community interface, poor ongoing monitoring and dosage adjustment	• No inferential statistics reported • Follow-up information incomplete • Self-report likely to overestimate compliance MMAT: 100 %
Reddy (2008) [39] To assess the extent to which evidence-based guidelines have influenced medical practice with respect to their experiences in depression assessment and management	Insights from surveys and interviews with GPs: • Little consistency among health professionals on how best to identify and manage depression • Few GPs asked patients about depression, regardless of patients’ depression score	• Wide distribution of guideline-related information was not effective in improving depression management • No agreement on appropriate time and provider for depression screening	• Published short report provides little detail regarding study design and quality appraisal • Study time frame not reported MMAT: n/a
Rushford (2007) [36] To assess patient recall of risk factor behaviour modifying intervention at discharge, 2, 4 and 12 months	Insights from study at 12 month follow-up: • CR referral is correlated with attendance • 8 % of women reported wanting more lifestyle advice	Limited advice provided on lifestyle (especially on diet & physical activity) to women who were obese or inactive. Older women less likely to recall receiving information Health staff need training in information delivery and communication skills	• Response rate 79 % • Good reasons for exclusion • Detailed assessment of recall on many areas of lifestyle • No details on how the initial patient education was conducted or its content MMAT: 100 %
Schrader (2005) [41] To evaluate the effect on depressive symptoms in cardiac patients of patient-specific advice to general practitioners regarding management of comorbid depression	In 2000–2001, in a randomized controlled trial: • The intervention had little effect on moderate to severe depression at 12 months • Telephone call to GP from psychiatrist led to a significant decrease in proportion of patients with moderate to severe depression • Multidisciplinary enhanced Primary Care case conference not effective (and difficult to implement)	Recommended screening of hospitalised cardiac patients for depression ansd providing targeted advice to their GPs	• No information on what management plans were actually delivered by GPs and no information on antidepressant prescription and service utilisation • Follow-up below 80 % (78.5 %) with differential non-reponse in younger separated/divorced patients and smokers • Allocation concealment unclear MMAT: 50 %
Schulz (2000) [37] To identify factors associated with and predicting attendance of post-MI patients at CR program	In 1993–1996, ~3.5 years post-MI: • 73.4 % referred to CR • Majority (72 %) of non-attenders were not referred to CR • Non-referral was significantly associated with non-attendance • Attendance significantly associated with referral	Being older, living farther away, living alone and not having private transport wre associated with CR non-attendance Referral to CR also predicted attendance	• 69 % response rate • Strengths and limitations of study well identified • Have not defined completion other than to offer second dropout rate of 36 % if attended 6 or fewer sessions MMAT: 75 %
Scott (2004) [31] To optimise care of patients with ACS and CHF through a QI intervention across two sectors (hospital and GP) of healthcare	In 2000–2002, at 3 months post-ACS, % of medications prescribed to patients: Aspirin: (baseline) 82 % (intervention) 89 % Aspirin continuation in those prescribed at discharge: (baseline) 84 % (intervention) 92 % βB continuation among those prescribed at discharge: (baseline) 76 % (intervention) 85 %	Implementing systems of decision support, targeted provider education & performance feedback, patient self-management and hospital-community integration improved patient care, particularly when directly controlled by individual clinicians (e.g., prescribing)	• Not possible to attribute specific process-of-care changes to specific QI initiatives within a multifaceted program • Only significant results reported MMAT: 50 %
Toms (2003) [34] To determine whether Phase II outpatient CR participants are more successful at achieving cardiac risk factor targets than non-participants at follow-up post-MI	In 2003, at 18–36 months post-MI: • Of 36 included CR non-participants (NP), 53 % not referred by doctor • CR participants less likely to have total cholesterol > 6.5 mmol/L • Fewer non-participants were receiving cholesterol lowering medication • In both groups, approximately 50 % did not achieve target total cholesterol (≤4.5 mmol/L) • CR participants more likely to be on lipid modifying treatment • More CR participants exercised regularly • Failure to achieve blood pressure and weight control similar in both groups • Small numbers continued smoking in both groups but insufficient sample size to assess statistical significance • More CR participants had returned to work (92 % vs. 78 %) but not statistically significant even after adjusting for age	Those attending CR had better long term outcomes, exercising more and more achieving the goal of a TC ≤6.5 mmol/l	• Participants resided within 40 km of Canberra therefore geography less of an issue • Used TC to assess lipids, not LDL • Highlight need for data collection • Low response rate (51 %) • Limitations of study recognised • Selection bias likely in those attending CR MMAT: 75 %
a. Wachtel (2008) [29] b. Wachtel (2008) [40] To determine assessment of lifestyle and behavioural risk factors in post-MI patients in hospital and at GP follow-up in a rural region of South Australia	In 2004–2005: • Population was 78 % male • One Aboriginal/Torres Strait Islander patient (2 %) • Majority of patients did not receive an intervention for risk factors • 5/11 (45 %) patient smokers received quit advice, one prescribed NRT • Higher proportion of patients received lifestyle interventions in GP practice than hospital setting, however, with the exception of smoking this accounted for 7 % of patients • 16/34 (47 %) patients had BMI assessed • 11 were overweight/obese of whom 2 (18 %) received weight loss advice	GPs generally increased prescribing of evidence based medications from time of discharge Major gap in CR and secondary prevention management of ACS patients in rural South Australia	• No documentation of special/additional services for ATSI population • Lifestyle and behavioural risk poorly documented except smoking status (76 %) and hypertension and diabetes (82 % and 78 %) • Low response rate MMAT: 50 %
a. Wai (2012) [20] b. Peterson (2012) [32] To improve the management of ACS at the point of hospital discharge, across the continuum of care	In 2009, at a median of 96-day- follow-up (range 49–204): • 48 % reported using 4 evidence-based medications (EBMs), with a significant decrease in anti-platelet agents, statins, β blockers and all 4 EBMs • 67 % recalled referral to CR of whom 33 % completed CR and 21 % were still attending CR • 731 GPs (47 % of patient-nominated GPs) participated in survey • 77 % received a discharge summary for patients with ACS at a median time of 3 days (0–41 days) after discharge • Of these 88 % contained a list of prescribed medications; 81 % included dose titration and duration of therapy and 55 % contained details of ongoing risk management • 65 % of GPs rated the quality of information as ‘very good’ to ‘excellent’ • 6 % increase in communication of ACS management plan to GP • 18 % increase in patients with documentated chest pain action plan	Targeted educational intervention can improve management of patients post-ACS Improvements evident in: • Evidence based prescribing • Communication between patient/carer 7 GP • Referrals to CR	• Accuracy of sample representation not documented • Based on medical record documentation and GP survey • Potential for Hawthorne effect • Low response rate of eligible GPs MMAT: 75 %

ACEi/ARB angiotensin-converting enzyme inhibitor/angiotensin-II receptor blocker, ACS acute coronary syndrome, βB beta-blockers, BMI body mass index, CABG coronary artery bypass grafting, CR cardiac rehabilitation, DAPT dual antiplatelet therapy, EBM evidence-based medication, GP general practitioner, GTN glycerol trinitrate, HDL high-density lipoprotein, LDL low-density lipoprotein, LLD lipid-lowering drugs, MMAT Mixed Methods Appraisal Tool, MI myocardial infarction, NSTEMI Non-ST elevation myocardial infarction, NZ New Zealand, PCI percutaneous coronary intervention, QI quality improvement, TC total cholesterol, UA unstable angina