This study will evaluate a complex intervention based on a patient management tool (PMT), combined with educational outreach to primary care doctors, nurses and other health workers, in the Brazilian city of Florianopolis. The intervention is aimed at improving the quality of primary health care and health outcomes, in adults with diabetes and cardiovascular disease (CVD). The effectiveness of the intervention will be assessed by randomly allocating 48 primary care clinics to receive the intervention or not, and comparing patient and clinic level endpoints that reflect the health and quality of care provided over the following year. About 11000 patients known to have been diagnosed with diabetes mellitus and 32000 with CVD (defined as having a diagnosis of hypertension, ischemic heart disease, heart failure or cerebrovascular disease) in participating clinics will be included in the study. About 7800 of them have diagnoses of both CVD and stroke. The primary endpoints will be 1. Number of participants in whom at least one of the following tests was recorded: body mass index, plasma glucose, serum cholesterol, electrocardiogram, and 2. in participants with a diagnosis of hypertension recorded previously, average systolic blood pressure recorded. Secondary endpoints will include the individual components of the composite scores, health measures (hospital admissions and deaths), and indicators of appropriate diagnosis of comorbid conditions such as depression. Eligible patients will be identified and outcomes measured using electronic medical records.

Diabetes mellitus and cardiovascular disease (diabetes and CVD) place a heavy and growing burden on people living in low and middle income countries. Many of them could be healthier if their disease was accurately diagnosed and correctly treated, but many are not. Doctors and nurses working in primary health care clinic are best placed people to diagnose and treat, especially where local clinics are near and free. But this raises two questions: 1. How to ensure that diabetes and CVD get the priority they need in overloaded clinics? 2. How to ensure rational evidence-based diagnosis and prescribing for diabetes and CVD? The investigators have developed a way of improving primary health care for people who have long term health conditions. It is a patient management tool (PMT), that is, a printed manual of flowcharts taking doctors and nurses from symptoms to diagnoses to treatments, tests or referrals, with advice on how to make decisions along the way about diagnoses, tests, treatments and referrals. They are prompted to think of other diseases and health problems that might be undetected or neglected. The package also includes a method of training known as outreach education. First trainers are trained, then trainers train groups of doctors and nurses at their workplaces, showing them how to use the guidelines, and using their own patients and clinical problems as examples. This outreach training is repeated several times in short sessions. The investigators' research in Africa has shown that this approach can be effective, cost effective, feasible and sustainable. It has been rolled out throughout South Africa and other African countries. But it has have not yet been shown to be effective for diabetes and CVD. The investigators have also not tried or evaluate it in Latin American countries, which have different health systems, and have many more doctors providing primary health care. Now co-investigators in the Brazilian city of Florianopolis have decided to put this educational package in place throughout the city, and have agreed to do so as a randomised controlled trial. This will clearly show whether PACK Brazil is effective, cost effective and feasible under Brazilian conditions. The core of the research will be a randomised controlled trial. 48 primary care clinics in the city will be randomly chosen either 1) to get the whole package of patient management tool plus training, or 2) only to get the patient management tool (which is expected to will make little difference without training). The investigators will compare patients in these two groups of clinics to see the effects of the training. They will use the clinics' electronic medical records to identify about 32000 adults diagnosed with diabetes and CVD. After the training starts they will follow these patients up for a year, and assess whether they are being appropriately treated and tested. The primary endpoints will be 1. Number of participants in whom at least one of the following tests was recorded: body mass index, plasma glucose, serum cholesterol, electrocardiogram, and 2. in participants with a diagnosis of hypertension recorded previously, average systolic blood pressure recorded. Secondary endpoints will include frequency of tests, the number who have each type of test, diastolic blood pressure in participants with hypertension, serum glucose levels in participants with diabetes, prescription of indicated treatments and treatment intensification, health measures (hospital admissions and deaths), and indicators of diagnosis and treatment of comorbid conditions such as depression. Eligible patients will be identified and outcomes measured using electronic medical records.

Inclusion Criteria:

• Age 35 years and over in July 2016
• Diagnosis of hypertension (ICD-10 code I10-I15), ischaemic heart disease (ICD-10 code I20-I25), heart failure (ICD-10 code I50), cerebrovascular disease (ICD-10 code I60-I69), or diabetes mellitus (ICD-10 code E10-E14) ever recorded since January 1st 2010, and attended a participating clinic for any reason during the first year of the trial

Exclusion Criteria:

• None