Community healthcare in Israel: quality indicators 2007-2009

Data for quality indicators are collected independently by each of the four health plans for their insured population. Each health plan maintains individual-level electronic data records that include medical appointments, procedures, laboratory test results and pharmacy claims. Quality indicators sent by each of the four health plans are aggregated and these values (numerators and denominators) are used to calculate national rates. Coordination, data compilation, validation and quality assessment and analysis of the national indicators are performed by the program directorate.

All Israeli residents are included in the data set. The quality indicators presented represent the time period 2007 through 2009. Electronic records were unavailable for a small percent of the population (< 1%). In addition, individuals whose membership in the plan was less than the full calendar year were excluded from the calculations. In 2009, for example, approximately 106,800 people or 1.4% of the insured population from the previous year switched healthcare plan [8] and did not have the required data for a complete measurement year. It is noteworthy that soldiers, largely comprised of individuals aged 18-21 years, are not covered by the health plans since their medical care is delivered by the military. Aside from these exceptions, the report includes the entirety of Israel's population, approximately seven million people.

Quality indicators were created with the consensus of the four health plans and Israeli medical associations. The HEDIS indicator set was used as a guide for the QICH indicator definitions. The QICH indicator set reflects a careful selection process that assessed the feasibility of production and the applicability and significance of the indicators for the healthcare system in Israel. Data were collected according to gender and age group. Six areas of healthcare - asthma, cancer screening, immunizations for older adults, child health, cardiovascular health and diabetes were used to assess the quality of community healthcare in Israel. Twenty-eight QICH indicators are presented for the years 2007-2009 (Figure 1). Identification of individuals with asthma and diabetes was according to pharmacy claims. Patients with cardiovascular disease were identified according to cardiac intervention - bypass surgery or coronary angioplasty. It should be noted, that these procedures were the only available diagnostic indicators of cardiovascular disease accessible to all health plans and therefore the number of patients included in this sub-population reflect a small percentage of individuals with the disease. Immunization status was established using pharmacy claims (i.e., vaccine purchase).

Data quality assessment as well as performance audits were carried out in three stages: by each health plan, by the QICH program directorate and by an accredited external auditor. Data checks that included logical tests for negative numbers, unit tests, and valid numerator and denominator entries as well as tests examining subgroups and changes over time were performed by each health plan internally and by the QICH program directorate. An external auditor conducted a procedural audit for each health plan as well as the for the QICH directorate. A system-oriented approach to examining the proper functioning and management of the program was performed and documented in comprehensive performance reports.

Rate differences were calculated as the absolute difference in percentage points between the rate in 2007 and 2009.

Rates of appropriate use of asthma control medication for individuals with persistent asthma increased over the three-year period and influenza immunization for this group increased to a rate of 40% by 2009 (absolute increase of 11%).

In 2009, the mammography rate for women over 51-74 years reached 68% and fecal occult blood test rate for colon cancer (excluding those who underwent a colonoscopy during the past five years) for individuals aged 50-74 years was 27%. Absolute rate increases of 5%-7% were observed for both types of cancer screening over the three-year measurement period.

Quality measures of primary prevention of cardiovascular disease, which included BMI (adults) and blood pressure documentation, increased to rates of 70% and 84%, respectively, for the 3.8 million Israeli adult target population aged 20-74 years. Rates of cholesterol documentation and control were 81% and 72%, respectively, for Israeli adults, ages 35-74 years.

Measures of appropriate treatment and effectiveness of treatment improved over the measurement period. For example, rates of use of medications, such as beta blockers and statins, for individuals following a coronary heart event were over 70%. The rate of appropriate control of cholesterol for cardiac heart patients was 72% in 2009; this represented an absolute change during the three-year period of +3.5%.

Over the three-year measurement period, positive changes in rates for cardiovascular care were observed for all aspects of treatment - prevention, care and effectiveness of care. In particular, substantial increases were observed for documentation rates of body mass index for adolescents and adults. For adolescents, rates of documentation of height and weight increased dramatically from 28% in 2007 to 61% in 2009, while for adults, these rates rose 26 percentage points over the same time period.

An indicator of quality of healthcare for primary prevention for children was examined. In 2009, 74% of children 1-year of age underwent anemia screening - an absolute increase of 7% from 2007.

Quality indicators for diabetes assessed both the provision and the effectiveness of care. Patients with diabetics mellitus received basic recommended care, such as glucose, blood pressure and cholesterol assessment (> 90%), BMI documentation (84%), kidney function testing (74%) and eye care documentation (64%). Only about half the population of diabetics received an influenza immunization in the winter season. Individuals with diabetes mellitus who attained adequate glycemic control (HbA1c ≤ 7%) decreased slightly over the measurement period (49% to 48%) and remained stable among those with poor glycemic control (13%). The percent of individuals with diabetes mellitus who reached pre-specified cholesterol and blood pressure targets showed an increase of 5.3% and 1.6% respectively.

The rates of influenza vaccination, a quality indicator of preventive care among Israelis 65+ years, increased over the measurement period from 52% to 57%.

The most dramatic increases in the QICH indicator set 2007-2009 were observed for documentation of anthropometric measures. Documentation rates for height and weight for the calculation of BMI among adolescents and adults increased by approximately 30% over this time period to rates of 61% and 70%, respectively. Rates in Israel are high relative to adult BMI screening rates in the United States (e.g., 35% for Medicaid and 41% for commercially insured populations) [12].

Moderate increases over time were observed for other primary prevention QICH measures, including immunizations for older adults, cancer screening, anemia screening and documentation of cardiovascular risk factors such as blood pressure and cholesterol. Comparison of these measures with those in other healthcare systems reveals wide variations in rates between countries. For example, the rate of influenza immunization for older adults in Israel in 2009 was 57%, similar to that of Ireland and Luxembourg and in contrast to higher rates in Mexico (88%) and France (71%) and lower rates in Finland (43%) and Hungary (32%) [13]. Mammography screening rates in Israel were similar to those of the OECD (68% versus 62%, respectively), although substantial variations exist between countries - high mammography rates in Finland (86%) and Ireland (78%) and low rates in France (47%) and Japan (24%) [14]. Anemia screening is not routinely conducted in most Western countries thereby preventing comparisons. Recently released guidelines of the American Academy of Pediatrics recommend universal screening for anemia for children 12 months of age to determine hemoglobin concentrations for the assessment of iron deficiency [15]. This step may generate future comparisons. Finally, documentation of cardiovascular risk factors, such as cholesterol and blood pressure levels, is essential for evaluating an individual's risk of heart disease [16]. Documentation rates of these factors have increased between 4%-15% from 2007 to 2009 and likely reflect the increased attention of the healthcare system to cardiovascular risks.

Rates of recommended care for three chronic conditions - asthma, cardiovascular disease and diabetes improved over the measurement period and are nearing those in other Western countries [12, 17]. For example, appropriate use of control medication for individuals with persistent asthma improved by four percentage points between 2007 and 2009; these rates remain lower than a similar HEDIS quality indicator (91%) [12].

For indicators assessing care of patients with cardiovascular disease, at least 70% of Israelis following a coronary intervention received the recommended treatment with statins, beta blockers or ACEI/ARBs. For patients following a coronary intervention, similarities between Israel and England were observed for treatment with beta blockers (Israel = 70%; England = 74%) but differed for treatment with ACEI/ARB (Israel = 69%; England = 89%) [17]. Notably the QICH indicator set defined coronary intervention as persons who underwent bypass surgery or coronary angioplasty in the past five years, while the QOF was based on patients who had a history of myocardial infarction since April 2003.

Treatment and care for diabetes in Israel has moved in a positive direction over the three year measurement period, although changes in rates for indicators of glycemic control were relatively small and even negative. Seven indicators of adherence to recommended treatment of diabetes mellitus were assessed. For the measurement period 2007-2009, higher rates (≥ 74%) of appropriate care were noted in four measures - documentation of BMI, kidney function, glycemic control and cholesterol, and relatively moderate rates (55%-74%) were observed in the remaining three areas - documentation of blood pressure and eye care and influenza immunization. The observed trends for adequate glycemic control (< 7%) are noteworthy. The results suggest that the practice of tight glycemic control may be changing among primary care physicians from a generalized HbA1c goal to an individualized target. Indeed, recent studies have shown that glycemic control for all patients with Type 2 diabetes may not provide the ultimate benefit [18]. One of the current goals of the QICH program is to revise this measure and determine sub-group-specific HbA1c goals. Decreases, albeit slight, in inadequate glycemic control (> 9%) indicate improvements in diabetes care. Further monitoring is necessary to identify long-term trends. Comparison with similar indicators and age groups from the HEDIS and England reveal moderate variations between countries and measures [12, 17]. For example, documentation of HbA1c among adults ranged from 92% in Israel, 97% in the England and 89% in the United States [12, 17]. Similarly, rates of microalbuminurea testing (kidney function) varied from 74% in Israel, 89% in the England and 84% in the United States [12, 17].

Improvements in quality of care have been noted previously in Israel [19]. In this recent study, the authors hypothesize that quality improvements may be a result of the feedback from the measurement process itself. This theory relates to the quality improvement maxim that "efforts to improve quality require efforts to measure it" [20]. To be sure, during the past decade of quality assessment of healthcare in Israel, health plans have implemented strategies for improving care and access to services as well as encouraged physician adherence to guidelines [21, 22]. The dramatic increases in documentation of anthropometric measures signify a targeted approach by health plans to document a neglected area of care and signal policy changes that have permeated all segments of the healthcare system. In light of these observations, expected trajectories of existing quality indicators are likely to attenuate over the next decade.

It should be noted, however, that incentives may account for lower rates of healthcare quality in certain areas in comparison to England's pay-for-performance system since the Israeli healthcare system is not a pay-for-performance system and physicians do not receive direct benefits for increased performance. In order to properly assess and compare the quality of Israel's healthcare system with that of a pay-for-service, outcome-based indicators and sufficient follow-up time are necessary.

Policy implications and future directions of this program mirror that of other international efforts [23, 24]. Quality indicators must undergo constant evaluation of whether they accurately capture the care that is being provided as well as their impact on health outcomes [24]. Examination of the adverse or beneficial changes in health outcomes is an essential tool in the evaluation process, since change alone in quality of care does not necessarily correspond to improvements in patient health. Composite indicators of health are also integral to disease management and future indicator sets should comprise such measures.

Assessment of quality indicators should comprise a more refined evaluation of subgroups and at-risk populations as well as changes over time in their quality care. In a recent report by Hussey et al. [10], similar disparities in socio-economic status (SES) in healthcare quality indicators were observed for four countries with different health systems and populations. The authors note that their study underscores the need to examine the impact of internal and external factors on the healthcare system (and health inequalities), such as changing healthcare patterns (internal) and targeted prevention programs (external). The proxy SES variable available to the QICH data set (data not shown) is of questionable utility and a proper assessment of the effect of SES on quality care in Israel is needed.

Finally, The QICH program has brought about substantial positive changes in the quality of healthcare data in Israel. An improved data platform that captures a larger share of the healthcare delivery system is necessary for the further assessment of continuity of care and other healthcare quality issues. Policymakers should promote the continued interface between the four Israeli health plans and support communication with other healthcare providers, such as for residential, early childhood and elderly care facilities. Interchange between hospital and primary care is vital to improved healthcare quality such as effectiveness and efficiency of care and care coordination. This type of improved framework requires policy changes and even incentives to allow for information exchange and data flow within the confines of privacy and confidentiality.

Israel is one of the only countries with a systematic and comprehensive evaluation program assessing the quality of the community healthcare at a national level. The QICH indicator set represents a unique and essential tool in healthcare. First, quality indicators are important to assess adherence of the community healthcare system to recommended guidelines that include preventive care. Increased use of preventive healthcare services is considered essential for reducing the clinical preventable burden as well as being cost effective [25]. Second, electronic health records - a mainstay of the medical system in Israel for over a decade, allow for the assessment of historical as well as current healthcare measurements. Third, all residents have unique identifying numbers that allow for linking external medical records, such as those for hospital procedures. Fourth, the program is a combined effort of all four health plan providers, who have, among other things, standardized disease registries and confirmed their commitment to treatment guidelines.

These quality indicators are not without limitations. Quality of community healthcare is anchored in its governing bodies and provider system and is susceptible to the constraints and shortcomings of each. International comparisons with similar measures are subject to the diversity of policy, culture and resources that directly and indirectly affect the structure, performance and outcome of healthcare [26]. In Israel, health plans are restricted by the availability of certain types of healthcare data, such as hospital discharge data and inpatient diagnoses. For example, patients with heart disease are identified using reimbursement codes for bypass surgery or coronary angioplasty even though this sub-population represents only about 10% of the patients with cardiovascular disease. The validity of the QICH indicator set must also be qualified. Apart from laboratory results, which produce objective and non-biased values, blood pressure measurements, for example, which require human assessment and documentation, vary between testers and may be subject to dishonest reporting. Lastly, regulatory restrictions of demographic information, including socio-economic status, limit the development of a complete assessment of quality across all categories of health and for various subgroups. Case-mix adjustments or stratification according to demographic or socio-economic characteristics may allow for a clearer understanding of the quality of health care in Israel in order to target gaps in care and identify and learn from successes.