Quality of Care

Quality of Care

Quality of care asks whether health services, as actually delivered, are likely to achieve valued health outcomes in line with current professional knowledge and patients' needs. It concerns more than the availability or volume of services: care can be accessible yet unsafe, technically sound yet poorly coordinated, or efficient on average yet inequitable. This page sets out the principal quality dimensions, explains how measures relate to care, and shows how an analyst can interpret a change without overstating its cause or value.

Describe the dimensions of good care

The World Health Organization (WHO) describes quality services as effective, safe, people-centred, timely, equitable, integrated and efficient. The Institute of Medicine's influential six aims cover safety, effectiveness, patient-centredness, timeliness, efficiency and equity; integration is explicit in WHO's list. These frameworks overlap substantially but should be named correctly rather than collapsed into a purported universal seven-item score.

Dimension	Practical question	What a narrow measure could miss
Effective	Does care based on good evidence improve relevant outcomes?	Delivering a recommended treatment may not mean it was appropriate for every patient.
Safe	Does care avoid preventable harm from care itself?	Recorded incidents may rise when reporting improves.
People-centred	Are preferences, circumstances and communication respected?	A satisfaction score may miss informed choice or unequal treatment.
Timely	Is needed care delivered without harmful delay?	A shorter wait for low-priority visits may displace urgent care.
Equitable	Is quality maintained across relevant population groups?	A favorable overall average can hide a widening gap.
Integrated	Do services and clinicians coordinate across a pathway?	A completed referral may not mean that follow-up occurred.
Efficient	Are resources used without avoidable waste?	A lower cost achieved by withholding beneficial care is not better quality.

Quality also depends on the condition, setting and population. An appropriate measure for elective surgery may be irrelevant to chronic disease follow-up. State whose perspective matters and whether the indicator concerns a facility, network or whole system.

Connect structures, processes and outcomes

The Donabedian classification separates what a service has, what it does, and what happens afterward. This helps identify a plausible improvement pathway, but the three categories are not interchangeable measures of success.

Measure type	Illustrative measure	Interpretation limit
Structure	Availability of trained staff or functioning equipment.	Capacity does not prove correct use.
Process	Eligible patients who receive a recommended intervention.	Recorded delivery does not prove benefit or fit for every patient.
Outcome	Complications, functioning, survival or patient-reported health.	Case mix, baseline risk and factors outside care also affect outcomes.

Patient experience may be measured alongside processes and outcomes, with the construct stated precisely. A facility can improve a process yet show little change in a rare outcome during a short measurement period. Conversely, a better outcome trend does not identify which process caused it. Choose a measure with a credible connection to what people value, valid data, a clear denominator and a practical opportunity to act on the result.

Calculate a process indicator transparently

Suppose a fictional clinic has 500 adults eligible for a specified annual review during a calendar year. Records show that 350 completed it. The observed completion rate is $350/500=0.70$, or 70%. After an improvement initiative, 420 of 500 eligible adults complete the review in a comparable year: $420/500=0.84$, or 84%. The observed change is $0.84-0.70=0.14$, a 14-percentage-point increase, equivalent to 70 additional documented reviews if the denominators and eligibility rules really are comparable.

Item	Spreadsheet expression	Result
Earlier completion rate	=350/500	70%.
Later completion rate	=420/500	84%.
Percentage-point change	=(420/500-350/500)*100	14 percentage points.
Extra recorded completions	=420-350	70 reviews.

This indicator measures documented completion among eligible people, not improved disease control or a 14% relative increase. The relative increase in the observed rate would instead be $(0.84-0.70)/0.70=0.20$, or 20%. Check whether more complete records, changes in who was deemed eligible, migration between clinics or a simultaneous policy change explain the difference. If the initiative targeted those least likely to attend, compare outcomes across relevant subgroups rather than only an overall rate.

Compare providers and periods fairly

A quality comparison needs a specified population, observation window, inclusion criteria and measure definition. Outcome comparisons often require risk adjustment for pre-existing severity and other baseline differences, but adjustment should not erase an inequity caused by care itself. Small denominators make rates unstable; a few events can produce large percentage swings. Report numerator, denominator and uncertainty where feasible.

Missing data and incentives also matter. A bonus tied to documented reviews may increase reporting or select patients easier to reach, so audit records and inspect balancing measures such as waiting times, complications and patient burden. Stratification can reveal disparities that a standardized overall number conceals. Changes in coding, referral boundaries or case mix can invalidate a seemingly simple before-and-after comparison.

Bring quality into economic decisions

Health economic evaluation should count relevant health gains, harms and resources, rather than using a quality score as a substitute for outcomes. A quality-improvement programme can require training, staff time, information systems and follow-up capacity; benefits might include avoided complications, better health-related quality of life and reduced downstream use. Specify the perspective and time horizon, compare with a feasible alternative, and avoid double-counting the same benefit as both a prevented event and its already embedded QALY gain.

Efficiency is a dimension of service quality, but lower spending is not automatically an improvement. An intervention can raise short-term costs and still offer good value if it produces sufficient health benefit; a cheap service can waste resources when it is ineffective. Equity concerns may also require separate distributional analysis rather than an average cost-effectiveness estimate alone.

Use measures as evidence, not as the whole concept

No single rate proves overall quality. Combine appropriately chosen measures with patient and clinician accounts, inspect unintended consequences, and test whether a change plausibly caused the observed results. Quality depends on clinical knowledge, organizational capacity and the experience of those receiving care; measurement should support improvement and accountability without encouraging attention only to what is easy to count.

Sources and further reading

• World Health Organization, “Quality of care,” for the seven service-quality characteristics: https://www.who.int/health-topics/quality-of-care
• World Health Organization, “Quality health services” fact sheet, for the definition and the role of quality in universal health coverage: https://www.who.int/news-room/fact-sheets/detail/quality-health-services
• Agency for Healthcare Research and Quality, “Types of Health Care Quality Measures,” for structure, process and outcome: https://www.ahrq.gov/talkingquality/measures/types.html
• Agency for Healthcare Research and Quality, “Six Domains of Healthcare Quality,” for the Institute of Medicine's six aims: https://www.ahrq.gov/talkingquality/measures/six-domains.html