How to conduct a clinical audit in six steps

The Care Quality Commission requires registered healthcare providers to monitor the quality and safety of their services and use the findings to protect the people who use them​[1,2]​. The ‘NHS long-term plan’ describes systematic methods of service improvement as essential to implementing best practice and adopting innovations​[3]​. Clinical audit is a useful quality improvement tool for monitoring interventions designed to address shortfalls in clinical care​[4]​. NHS England describes a clinical audit as “a way to find out if healthcare is being provided in line with standards” and that it “lets care providers and patients know where their service is doing well, and where there could be improvements”​[5]​. Whatever sector a pharmacy professional works in, they are accountable for monitoring practice and assessing and minimising the risks of the care provided​[6]​. Arguably, more change and improvement in healthcare comes from clinical audits than from research (because more audits than research are conducted), so the ability to conduct a good clinical audit is a valuable skill​[7]​.

A pharmacist’s employer may have contractual obligations to participate in certain audits. For example, the National Clinical Audit and Patient Outcomes Programme consists of more than 30 national audits related to some of the most common health conditions. The most recent National Diabetes Audit of National Institute for Health and Care Excellence (NICE) standards for diabetes care highlighted the negative impact of the pandemic on glucose and blood-pressure control across the country and, importantly, demonstrated greater than usual geographical variation in care processes​[8]​.

An employer may also have an internal audit programme. Many of these clinical audits will have arisen from previous audits, patient and public engagement activities, organisational clinical priorities, the clinical risk register, commissioned targets, adverse incidents, near-misses or complaints​[9]​. Many NICE guidelines have a corresponding NICE clinical audit tool. This article describes the steps to follow to choose and conduct a meaningful audit that improves the quality of care.

Step 1: Decide what to audit

To devise a local clinical audit topic, consider how the care provided meets the relevant standards for safety, efficacy, providing a positive patient experience, and regulatory requirements​[9]​. Is there any evidence of a quality problem? Are there potential economic or efficiency benefits to improving a care process because it is done so often? For example, is medicine reconciliation always completed within seven days of hospital discharge, or are prescriptions completely consistent with the organisation’s standards for prescribing​[10]​?

The Donabedian model for measuring quality care is an essential component of quality improvement in the NHS (see Table )​[11]​. According to this model, outcome measures are the ultimate validator of the effectiveness and quality of healthcare but can sometimes be difficult to obtain or have a long lag time. Process measures are, in fact, equally important, because they describe whether standards of care have been ‘properly applied’; without them, it is not possible to show a causal link between process and outcomes​[11]​.

There are three prerequisites for any clinical audit:

• Evidence-based clinical standard(s);
• Audit proforma derived from the standard(s) — a proforma is a data collection tool or checklist to record the outcome variables measured in the audit; using one ensures that the same set of data is collected in the same way for each subject (protecting data quality) and improves the efficiency of data collection;
• A clearly defined population of patients (or a sample from the population of interest) whose care will be measured using the proforma, usually retrospectively and from their healthcare record, although it can also be done prospectively by the clinician or observer​[4]​.

If any of those prerequisites are missing, the topic cannot be investigated using clinical audit. Either research methods or other quality-improvement tools will be more suitable (see Box). The NHS Health Research Authority has developed guidance and a decision tool to help determine what research method the question being investigated needs​[12]​.

Step 2: Develop a project plan

A poorly designed clinical audit is unlikely to achieve its aims and is a waste of time and resources​[13]​. The Healthcare Quality Improvement Partnership (HQIP) ‘Best practice in clinical audit’ guide sets out best practice for conducting local clinical audits​[14]​.

If a pharmacist’s organisation does not have its own checklist for conducting a clinical audit, NICE and HQIP have published checklists to support well-designed and well-conducted audits​[15,16]​.

Clinical audits should be designed and carried out with or by the people who provide the care or service being audited​[4]​. Everyone who is involved in delivering the service should be informed about the audit from the beginning​[14]​. Consider including anyone who may be involved in acting on the findings, such as clinicians in adjacent services, service users and carers. Desired standards are then embedded and awareness is raised among those delivering the care​[4]​.

The project needs clear aims and objectives​[14]​. Objectives should be SMARTER (Specific, Measurable, Achievable, Relevant, Time-based, Evaluated and Resourced) for the best chance of success​[17]​. The methods must be clearly documented so that they are reproducible​[14]​.

It is a common misconception that only research projects require a consideration of ethics​[16]​. Take advice from a specialist, such as the local ethics advisory service or independent pharmacists association, if the proposed project:

• Places additional burden on service users over and above routine care (e.g. they are required to have longer appointments for data collection);
• Allocates any interventions differently among groups of staff or service users;
• Involves the use of interventions that are not usual clinical care or are untested;
• Provides no direct benefit to patients or patient care (e.g. designed solely to control or reduce costs)​[13]​.

Include data analysis and appropriate sample size at the planning stage of the project​[14]​. Free sample size calculators (with guidance for use) are available online, such as the Compliance Master sampling tool, which lends itself to binary (i.e. yes/no) audit data and the ClinCalc sample size calculator, which can handle averages of continuous data (e.g. mean blood pressure reduction) in addition to binary data. Frequencies and percentages are often all that is needed from clinical audit data, but choose statistics that give the clearest possible picture of performance against the relevant standard(s). In addition, consider the granularity required for reporting — for example, individual consultants may wish to use their data as part of revalidation​[14]​.

Use and transfer of service users’ identifiable data must be managed in line with current information governance legislation and the eight Caldicott Principles​[14,18]​. This is applicable at all stages of the project, including sample identification, data collection and analysis​[14]​. Seek advice from a Caldicott Guardian if a novel or difficult decision is required​[19]​. Only data required to measure performance against the audit standard should be collected​[14]​.

The data extraction process and audit proforma should be piloted before collecting data at scale​[20]​. The proforma should specify the data sources for each standard, using data sources that yield the most quality data for the least amount of effort. Decide in the design phase how to approach issues likely to arise in the data sources, such as illegible handwriting in paper record; technological expertise required for digital systems; missing data and poor response rates for postal questionnaires​[20]​. If there is more than one data collector, build decision support into the audit proforma to reduce data variability.

Step 3: Gather data

Collect the data using the audit proforma. It is a good idea to input and sense-check the data as soon as possible after collection. Data collection mistakes are easier to correct if they are spotted quickly​[14]​.

If a serious incident is detected during data collection that could have an important effect on a service user, the registered provider is obliged to disclose it, support the individual and take steps to ensure that the system or process involved has been improved​[13,21]​.

Step 4: Analyse data against standards

Compare the results with the audit standard(s). Careful data analysis is necessary to ensure that the correct conclusions are drawn. The HQIP guide to analysing quality improvement and assurance data is an excellent resource for anyone new to this aspect​[22]​.

Step 5: Identify areas that require improvement and implement improvement strategies

If the audit finds that a service does not comply with the audit standard(s), look for the causes of the non-compliance​[14]​. Several quality improvement techniques might be helpful here:

• Process mapping;
• The ‘five whys’;
• Cause and effect (fishbone) diagrams​[4,23,24]​.

If a new perspective on the problem is needed, the Q community has a toolkit of creative approaches to problem solving that is available to non-members online​[25]​.

Once the causes of non-compliance have been identified, an improvement strategy can be developed. An improvement strategy can include more than one planned improvement and might involve other quality improvement tools, such as performance benchmarking, communication tools, decision trees, new technology or Lean Six Sigma​[14,26]​. ‘Plan Do Study Act’ (PDSA) cycles are helpful to refine the intervention before it is implemented​[26]​. A large organisation will have clinical audit practitioners and offer training for staff in quality improvement methods​[14]​. A self-directed e-learning package on quality improvement for healthcare professionals is available from the Healthcare Quality Improvement Partnership​[27]​.

Step 6: Re-audit

Re-audit should take place after implementing the intervention and, ideally, using the same audit proforma​[14]​. The cycle is not complete until improvement is seen, so it may be necessary to repeat the cycle more than once. This is sometimes referred to as an audit spiral. It is good practice to repeat the audit to confirm that improvement has been sustained​[4,28]​.

After the audit

The audit findings, required actions for quality improvement, and challenges — and how they were overcome — should be shared with the entire relevant workforce to promote learning​[4,28]​. This may include colleagues in partner organisations, commissioners and other professional groups​[28]​. Consider also publishing your work and sharing your findings with a local or national community of practice, such as clinical audit networks, academic health science networks, primary care networks, the Q community or other similar groups, to enable the sharing of experiences, knowledge and expertise within local and wider health systems​[29,30]​.

Summary

Clinical audit is a valuable tool to improve the quality of care a service provides. Following the six steps outlined in this article will help to deliver an effective and successful clinical audit.