11.1.3.3. Quality in health care

 

After a period of cost containment and efforts to improve efficiency, the recent years have seen governments increasingly focusing attention on effectiveness in the form of quality of care. For instance, evidence from the US provides some indication of how quality can be measured, identifies significant variations in quality and shows that improving quality is very difficult (and that efforts have generally been unsuccessful) (Brook et al, 2000). Moreover, due to differences in the way quality is measured and defined, there is little evidence available comparing quality of care across countries. Quality encompasses a multitude of dimensions, including effectiveness, access, responsiveness, patient-centeredness, technical competence, equity, appropriateness, availability, respect, timeliness, patient/care experience, choice/availability of information, continuity and prevention/early detection. Also integral in quality of care is patient safety, which is discussed here below. The lack of a common systematic framework, or even consensus, on how to define quality of care, is largely due to the wide diversity in the language used to describe the concept (Blumenthal, 1996; Brook et al, 1996; Evans et al, 2001; Shaw and Kalo, 2002). However, a recent OECD working paper sets out a proposal for such a framework (Kelley and Hurst 2006).

 

Nevertheless, since the mid-1960s the quality of healthcare has been measured in terms of three primary components: structures of care, processes of care and outcomes of care (Donabedian, 1980). Structure refers to a health care provider’s or facility’s capacity to provide high quality care. Structural variables encompass the level, mix, education and training of staff, and the characteristics of the facilities in relation to the characteristics of patients or residents, such as demographics, payer mix and case mix. It also includes issues such as the safety and appropriateness of the environment and the availability of updated health technologies. Process measures the services actually provided or administered. Deficiencies in processes of care can be described as overuse or underuse of care, or poor technical performance. Outcomes represent changes in health status due to provided or not provided care. In quality assessment, two types of outcomes are generally measured: subjective/self-assessed and objective/clinical outcomes.

 

Moreover, distinctions can be made between organizational quality assessment models and clinical quality assessment schemes (Øvretveit, 2001). The former are directed at evaluating the organization of care and cover instruments such as accreditation, certification and external and self-assessment schemes, whereas the latter include approaches such as peer review/visitation, clinical audit, quality circles and medical specialty.

 

Quality indicators are necessary to guide clinical quality assessment although, because of difficulties in developing appropriate indicators, few countries make use of them. For example, in Sweden, voluntary quality registers were developed for health professionals in order to disseminate good medical practice, provide comparative performance data and encourage continuous quality improvements (Rehnqvist, 2002 as cited in Smith, 2004). Registers are based on clinical specialties - e.g. cataracts surgery - and managed by teams in university hospitals. Despite being voluntary, a large proportion of physicians (about 70%) participate.

 

Publicizing quality

 

Some countries have gone further, developing quality indicators and reporting them to the public. Performance data released to the public arguably serves to: increase public accountability of healthcare organizations, professionals and managers; maintain standards and improve quality (Marshall et al, 2003). There are several assumptions underlying this argument that may or may not be met in practice:

Patients make rational choices

Information asymmetry between purchasers/providers/patients is minimal or, at least, not highly influential

Purchasers contract on quality (including safety) not just price

Providers respond to a reduction in patients by improving quality instead of cream-skimming

There is a direct relationship between strategies of care and outcomes

 

Evidence that reporting performance to the public improves quality is mixed. In some countries patients do not routinely use performance data to choose providers (e.g. Denmark and some states in the US), while early evidence from the US, Denmark and New Zealand suggests providers respond to publicly reported performance data. Recent evidence from the US shows that some purchasers are using information on performance to incentivize providers when negotiating contracts (Goldfarb et al, 2003; Epstein et al, 2004; Mainz et al, 2004). Observational evidence from the US shows that those states with public reporting systems have experienced a faster decline in cardiac mortality compared to the states without public reporting systems. The evaluation of the Danish indicator project established in 2000 showed that the publication of performance data is professionally accepted when coupled to audit comments; it can enhance improvement activities; has clear effects on priority setting; yet lacks of impact on patient empowerment (Mainz and Bartels, 2006).

 

International efforts

 

In 2001 the OECD launched the Health Care Quality Indicator Project to track quality of healthcare across countries. The long-term objective is to develop a set of healthcare quality indicators that can be reliably reported across countries using comparable data. A set of comparable data across 23 countries was developed, and consensus recommendations of an international expert group provided indicators for five priority areas, including: cardiac care, diabetes, mental health, patient safety and primary care/prevention. Indicators included in the initial OECD were selected on the basis of scientific validity and reliability in addition to three general criteria: impact on health, policy importance and susceptibility to influence by the health system (Mattke, Kelley et al, 2006). Among the indicators there were breast cancer survival, mammography and cervical cancer screening, waiting time for femur fracture surgery and asthma mortality rate.

 

Looking at breast cancer and cervical cancer screening, data are so far available for 18 of 23 European countries in the OECD. For these countries, it appears that high rates of mammography (greater than 80% of women aged 50-69 screened) can be found in Finland, Sweden, the Netherlands and Norway, with lowest levels in Czech Republic, Hungary, Italy, and Switzerland (50% or less). For cervical cancer screening, rates are lower than for mammography, with higher levels (greater than 70% of women aged 20-69 screened) seen in France, Iceland, Ireland, Norway, Poland and Sweden, and with lower rates (less than 50%) in the Czech Republic, Denmark, Hungary and Italy. Clear improvements can be made in all countries to increase the rate of screening among the target populations (Mattke, Kelley et al, 2006).

 

In addition, the International Organization of Standardization – a network of national standards institutes - publishes guidelines on numerous medical-related procedures ranging from medical equipment, sterilization and disinfectants, to more general guidelines for process improvements in health facilities. At European level, the European Foundation for Quality Management provides training and guidance for numerous quality improvement initiatives covering all sectors.

 

Appropriateness of care

 

Appropriateness of care is a dimension of quality that refers to the extent to which the treatment corresponds to the needs of the patient. Strategies to ensure appropriateness of care relate to clinical quality assessment schemes such as clinical practice guidelines, quality indicators and information systems, patient surveys, clinical governance and audit processes.

 

Clinical practice guidelines are specific criteria for how and when particular tests and treatments should be used for specific diseases and conditions based on the best available evidence. They may reduce disparities in treatment across physicians as well as control spending (Walley and Mossialos, 2004). While it is difficult to monitor and enforce these guidelines, financial incentives (or disincentives) coupled with educational efforts may improve compliance. Many European countries are developing guidelines and sharing information, as in the EU funded AGREE project (Burgers et al, 2004). While some countries already have systems in place - e.g. the Czech Republic, Finland, France, Spain, the Netherlands and the UK - others are beginning to introduce guidelines e.g. Austria, Belgium, Cyprus, Estonia, Latvia, and Poland.

 

The effectiveness of clinical guidelines on the quality of care is uncertain (Gundersen, 2000). While some studies demonstrate minimal effect on physician prescribing behaviour (Hetlevik et al, 2000), others suggest that well-designed and consistently implemented guidelines can help to deliver “best practice” (Garfield and Garfield, 2000; Richman and Lancaster, 2000; Perleth et al, 2001). While most guidelines seek to improve the quality of care, others are designed with the explicit objective of cost-containment and are unlikely to be acceptable because of ethical and legal implications (Carter et al, 1995; Cheah, 1998). Also, if clinical guidelines are not legislated, as in Finland, they may not be as successful in improving quality of care. In France, there has been poor compliance with prescribing guidelines for many reasons: the volume of guidelines, lack of information systems and limited capacity for monitoring, together with the physicians concern that following the guidelines could negatively affect the quality of care being delivered (Durieux et al, 2000).

 

Regulating prescribing patterns is another method used to ensure a suitable level of appropriateness of care. Prescribing patterns differ significantly across countries. For instance, only 62.9% of consultations result in prescriptions in the Netherlands, compared to 94.5% in Italy. Various approaches have been made to monitor prescribing quality, such as the use of a Medical Appropriateness Index which assesses prescribing suitability (Mossialos et al, 2004). In the UK, prescribing data are used to provide doctors with reliable and regular information on their current prescribing in an attempt to encourage more effective and economical prescribing.

 

Pharmacies play a role in ensuring appropriateness of pharmaceutical care. Hospital pharmacies’ functions have radically changed since the 1970s and now must support the safe, effective and economic use of medicines in hospitals in accordance with government rules and budgetary requirements. Hospital pharmacies increasingly provide products to meet the individual patient’s need, thus require an increased collaboration between hospital pharmacists, prescribers, nurses, dieticians, biochemists and laboratory scientist. Medical information and clinical pharmacy services are needed in the hospital to service outpatient care. The role of hospital pharmacists has also expanded to include working as clinical pharmacists at ward level, along with their traditional responsibilities of drug preparation and verification. In addition, specialized databases and medicine information services based in hospitals have been developed to facilitate drug treatment decision-making by clinicians (Taggiasco et al, 1992).

 

Provider payment methods and quality of care

 

Direct financial incentives to improve quality of service provision are used in many countries. Quality specifications in a payment contract can be structure, process or outcome oriented, thereby differing in what they measure, but all enacted to meet the goal of quality provision. One of the most ambitious initiatives to promote quality in general practice was the 2004 GP contract in the UK. The contract rewards GPs financially for meeting 146 indicators drawn from four domains (and a fifth bonus indicator on 'access'). The four domains are: clinical standards (covering major disease categories), organizational standards (related to information), patient experiences and additional services. Each indicator is weighed, contributing to an overall maximum quality score for each practice of 1,050 points, with scores of individual practices made publicly available.

 

Structurally-oriented specifications refer to quality measurement mechanisms such as data collection systems or internal quality management mechanisms (Shaw, 2003). In Germany, health insurance funds impose data and documentation collection efforts on German hospitals to monitor performance on specific diseases and interventions. In the UK, the GP contract includes patient documentation requirements. Quality indicators related to process seek to monitor patient outcomes through specific protocols and guidelines. These may be vaccination rates or specific disease treatments (as for coronary heart disease in the UK GP contract). Process-related indicators may also refer to a specified volume of services. Similarly, outcomes targets make use of guidelines, although they are based on results of treatment as opposed to the treatment levels themselves. Examples of outcome target include the setting of patient blood pressure and cholesterol targets in the UK GP contract, while a French hospital contract set that within two years the rate of nosocomial infections should be reduced by 30%. Linking sanctions or financial incentives to outcomes is highly complex. For example, adjustment for case mix is needed to identify highly complex patients who may be skewing health outcome data (Mossialos et al, 2007).

 

There are a number of concerns with the design and impact of pay-for-performance schemes. For example, it is unclear whether financial incentives are sufficient to motivate and support the necessary structural investment and behavioural change, and if they will raise quality only in those already performing well. Further challenges include the cost of acquiring information technology, a multiplicity of programmes and guidelines and the difficulty of data collection. Perhaps as a result of these difficulties there has been little evidence to support the effectiveness of paying for quality in healthcare (Rosenthal and Frank, 2006). However, these programmes have the potential to increase the physician’s use of evidence-based clinical guidelines, administrative and clinical best practices, information systems and access to appropriate and timely care (American Academy of Family Physicians, 2006). Finally, as paying for quality will entail additional administrative costs to the system, these programmes should be judged based on some notion of value or cost-effectiveness relative to alternative interventions for improving healthcare quality (Rosenthal et al, 2004; Campbell, Reves et al, 2007).

 

There is inconclusive evidence on the link between targeted financial incentives and the behaviour of individual doctors, particularly in relation to quality improvements. Methodological problems persist with attributing complex behavioural changes to particular interventions and with assessing the spillover effects of interventions onto behaviours other than those incentivized (Marshall and Harrison 2005). Other challenges with assessing the effects of financial incentives on physician behaviour include observations that beyond the economic rewards of financial incentives, doctors are motivated by other goals e.g. improving chronic disease management (Spooner et al, 2001), and may have a targeted income beyond which they are no longer motivated by financial incentives i.e. no linear relationship between incentives and impact (Rizzo and Blumenthal, 1996). Despite these challenges, there are some important lessons regarding financial incentives that can be taken from the above studies: they are more effective if they are owned by their target audience and aligned to professional values; they should be focused more on technical aspects and less on indeterminate aspects of professional practice; moreover, professional motivation is more likely to be damaged by overly bureaucratic schemes (Mossialos et al 2007). Thus, it would be inappropriate to link financial rewards to complex diagnostic processes or to the psychosocial aspects of care provision.

 

Patient and public experiences

 

Public and patient surveys shed light on the level of responsiveness within the health system, and overall levels of satisfaction. Moreover, public opinion surveys such as the Eurobarometer surveys coordinated by the European Commission are useful for comparing differences in opinions across countries and time. They do, however, suffer from limitations relating to difficulties in generalizing results to the population (due to survey design), and potential bias arising from socio-cultural factors that may influence reporting styles. Results from these surveys are also inherently subjective, while cross-country differences may reflect, among other things, differences in expectations, economic climate, or political dissatisfaction.

 

While one survey on perceived quality of health and social care suggests that there is some variability across the EU (Anderson, 2004), assessments can be divided broadly into the EU15 countries having higher scores than the new Member States of 2004, with the highest levels reported in Austria. There are, however, some exceptions. Several EU15 countries, notably Portugal and Greece, but also Italy and Ireland, are below the mean score for all countries, while the ratings of people from Malta and Cyprus put their countries among the top half for health and social services. The highest ratings of perceived quality of health and also social services can be seen in Austria. It is difficult to glean any conclusions based on this survey data because of difficulties in comparing such subjective indicators of quality; indeed many argue that within the European region, cultural differences may be driving a large part of the variation.

 

Table 11.2. Perceived quality of public services

 

Patient safety

 

Patient safety is increasingly recognized as integral to ensuring overall quality, and is only slowly being prioritized in Europe where few countries have formal systems in place. In the UK, the National Patient Safety Agency was established in 2001 to improve patient safety by reducing the risk of harm through medical errors via the promotion of a culture of learning from adverse events. This followed a Department of Health report, An Organization with a Memory, highlighting the human and financial costs of medical errors. The Agency operates an annual national system of reporting adverse events. Data are held anonymously and disseminated across the country to look retrospectively at incidents in order to determine what happened, how and why. The Agency also offers resources such as learning tools for NHS staff to help raise awareness on patient safety issues. Denmark is another country with a formal system for ensuring patient safety.

 

Patient safety can be improved through various efforts. Education is one way - through formal pre-and postgraduate training - and in experiential learning including continuing professional development (CPD). Modern professional training courses in patient safety should demonstrate a culture of safety, one which: acknowledges and learns from mistakes; encourages all to improve quality and safety; and is supportive of all personnel to protect patient safety, regardless of the rank (Howe, 2006). Rules and regulations can also be used, for example through evidence based practice (EBP) (see also “Appropriateness of care” above). Some countries, for instance the UK through the National Institute of Health and Clinical Excellence (NICE), are increasingly incorporating EBP into their guidelines. At international level, the World Alliance for Patient Safety was formed in 2002 and passed a resolution urging the World Health Organization to develop global clinical norms and standards. Nevertheless, it is important to highlight that even when guidelines or protocols are available, healthcare professionals have been shown to deviate from them. It is therefore crucial that rules designed to influence behaviour must be understood and accepted by those expected to use them (Claridge et al, 2006).

 

Encouraging teamwork and communication on adverse events is also used as a manner of promoting a culture of patient safety (McCarthy and Blumenthal, 2006). For instance, the surgical oncology and general surgery intensive-care units at Johns Hopkins Hospital implemented a safety scheme that had staff follow eight steps to reduce patient safety concerns. This resulted in a one-day decrease in average patient length-of-stay and 43 less catheter-related infections annually, saving eight lives. Quick response and the recognition of early warning signs also play key roles in preventing the potential long-term negative impacts of medical errors. An example comes from the Missouri Baptist Medical Center in St. Louis, Missouri which initiated a rapid response system to early signs of declining patient health to avoid acute crises. This resulted in a 15% reduction in cardiac arrests for the 489-bed hospital and a 3.95% decline in the hospital’s mortality rate (McCarthy and Blumenthal, 2006). These successes demonstrate how improving teamwork across hospital staff and the implementation of a systematic approach, can save both lives and money.

 

Medical errors

 

Medical errors present a significant, although largely preventable barrier to high quality health care. They have therefore become a point of much attention for policy makers, medical professional and patients. Medical errors result in approximately 44,000-98,000 unnecessary deaths and over one million injuries per year in the US (American Hospital Association, 1999). In Australia they account for about 18,000 unnecessary deaths per year in addition to 50,000 patients becoming disabled, with human error accounting for most adverse events (81.8%) (Wilson et al, 1995). Less research has been conducted at European level, although estimates suggest that errors occur in about 10% of hospitalizations and that about half of these could be prevented (Mossialos et al, 2007).

 

In the US, fatalities from prescription errors were found to have increased by 243% from 1993 to 1998, outpacing almost any other cause of death, and also progressing faster than the increase in prescriptions (Phillips and Bredder, 2002). One study highlighted errors in prescribing medications as the most common mistake among family physicians (Dovey et al, 2003), while another, in an American teaching hospital, reported four errors per 1,000 medication orders, 70% of which had the potential to be seriously harmful (Lesar et al, 1997). Preventable adverse reactions to drugs are claimed to be the single leading cause of hospitalization in the US, where 2-7% of hospitalized patients have avoidable adverse drug events and, consequently, have hospitals stays 8 to 12 days longer than they should (Kohn et al, 2000). UK studies show similar results, with one report of a 49% error rate in the administration of intravenous drugs (Taxis and Barber, 2003). Patients who face multiple interventions and have more serious conditions, causing them to remain longer in hospital, are more likely to suffer consequences resulting from a medical error. Otherwise, all patients regardless of sex, age and level of co-morbidities face medical error risks (Weingart et al, 2000). Medical errors have been attributed to a number of causes, among which: administrative and investigation failures, simple ignorance, lapses in treatment delivery, miscommunication, complications in payment systems, etc. (Dovey et al, 2002). Not just technical incompetence among healthcare professionals, but medical errors also arise from poor design of health care delivery processes (Kohn et al, 2000). Without robust information systems to track medical error occurrences and difficulties in documenting their occurrence, such as universal underreporting, assessing the actual size of medical errors remains difficult.

 

Not only do adverse events represent a concern for both patient safety and the finances of the healthcare system, as they lead to greater medical complications or even deaths. In addition, errors lead to intangible costs, such as diminished trust in health systems, medical professionals and individual hospitals, as well as the physical and psychological effects of prolonged hospital stays and/or disability levels (Kohn et al, 2000).

 

A cross-country study on public perceptions of medical errors in Europe was released in January 2006 within a Eurobarometer survey (European Commission, 2006). The survey included all then EU25 Member States as well as Bulgaria and Romania (then still accession countries) and Croatia and Turkey. Findings illustrated a wide variety in risk perceptions associated to medical errors and experiences pertaining to mistakes. Some 78% of EU citizens found medical errors to be an important problem in their country, but country-level data found ranges from 97% in Italy to 48% in Finland. When asked if they are worried about suffering a serious medical error, Austrians, Swedes and Dutch respondents were the least worried. Women, older age and lower education levels tend to be associated with a higher likelihood of perceiving medical errors as an important problem. Those who have actually experienced a medical error or who have had a family member in such a situation tend to be more likely to view the problem as important, and to be more prone to worrying about suffering a medical error. Hospital incidents are perceived to occur more often than medicine-related errors. When looking at trust levels for individual professional groups, most Europeans have confidence that their doctor, medical staff and dentist will not make any mistake while treating them. Respondents were most confident in their dentist, with the EU25 average being 74%. There are varying degrees of confidence amongst the EU25 for doctors and other medical professionals. Finnish citizens have the highest level of confidence in both categories (as well as with dentists), while Greece, Poland and Latvia have the lowest.

 

In brief, the Eurobarometer survey demonstrates that medical errors are perceived by Europeans as a problem and citizens are well aware of their occurrence, with 78% having read or heard about them. While the majority of respondents expressed confidence in health professionals, a sizeable sample did not. Although patients believe the health system is responsible for avoiding medical errors, the patient is also seen as having a role in decreasing their own likelihood of experiencing a medical error.

 

Satisfaction with the health system

 

Patient and public satisfaction with healthcare is measured through population or patient surveys and opinion polls. Comparative data across countries tend to be based on public rather than patient experiences. The European Social Survey from the years 2002, 2004 and 2006 provides a level of satisfaction with the health systems ranked from 0 (least satisfied) to 8 (most satisfied). Highest satisfaction levels in 2006 can be seen in Belgium, Finland, Austria and Switzerland (6.5 and over), the lowest in Bulgaria, Portugal and Hungary (less than 3.5) (See Figure 11.1). In most countries, there has been an increase in general satisfaction or little change over the years, though some have seen a decline (e.g. in Slovakia, Hungary, Germany).

 

Eurobarometer surveys also provide an indication of public satisfaction levels across Europe, but the most recent survey included only EU Member States prior to 2004 (Eurobarometer 2002). Among these 15 countries, Finland and Austria had the highest proportion of individuals agreeing with the statement “the health system runs quite well” (as also shown in the European Social Survey), while Greece, Italy and the UK had the lowest. Furthermore, only Italy and Spain saw an increase in the percentage of individuals who held a positive view of the health system from 1996-2002. The remaining countries all saw a decline in satisfaction over the period, most significantly in the Netherlands (from 70% to 45%) and Denmark (90% to 50%). The survey also questioned individuals about the extent to which doctors spent adequate time with them in general, and specifically regarding preventive health and lifestyles. When asked whether they agree or not that 'doctors do not spend enough time with you when you go to them', respondents have the opportunity to express how satisfied they are with the quality of the service given by their doctor. The question thus targets a specific and common service, and is taken as indicative of perceived quality of physician care. It is therefore very different from questions about satisfaction towards the healthcare system as a whole.

 

Figure 11.1. Overall satisfaction with the health system (scale of 1-8)