Improved Quality Diabetes Indicators, Outcomes and Costs
Improved Quality Diabetes Indicators, Outcomes and Costs
Our study has demonstrated that over a 6-year-period, improvement in glycemic and cholesterol control was associated with significant decreases in hospitalization days, mortality and direct medical costs. ED visits failed to demonstrate a similar association.
Results in Context. These results may be explained by the fact that improved glycemic and cholesterol control were important elements in the organization's efforts to improve quality of care, through multidisciplinary health promotion programs created within the framework of a performance monitoring system that implemented the organization's "call for action".
In the early 1990s, the Institute of Medicine (IOM) defined quality as the "degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge". Quality of care can then be evaluated on the basis of structure, process, or outcome. When used appropriately, both process and outcome measures can provide valid information about quality objectives.
The decision to choose hospitalizations, ED visits and mortality as health outcomes (the dependent variables) was based on data availability. Data on diabetes complications and long-term trends in diabetic patient satisfaction remain unavailable. It has been long claimed that the use of outcome measures is insufficient for quality of care assessment since most differences between patients receiving the same treatment result from factors associated with patients' characteristics – which are outside the control of health care providers. Following our experience, it is believed that beyond the differences in personal characteristics, differences in care provided to MHS members tend to be marginal. However, differences in personal characteristics (gender, age and socio-economic rank) were adjusted in all the statistical models. It is suggested that further research will be conducted with a control for patient characteristics such as self-efficacy, self-capability to manage the disease, compliance with medical regiment, and so forth.
Data obtained from the Israeli National Program for Quality Indicators (NPQI) in Community Healthcare indicated a 0.25% annual increase in diabetes prevalence from 2004 to 2010. National census data indicated that diabetes-related mortality rates, adjusted for ethnicity and gender, have decreased by 24% from 2004 to 2010, most probably due to better control of intermediate outcome measures and better care for diabetes complications. Although data on diabetes-related hospitalizations are not available, crude national figures indicate a steady increase in hospitalization days between 2000 and 2010.
It is worth mentioning that the NPQI, instituted in 2004, contributed to each of the four Israeli health plans, including MHS, for development of quality improving infrastructures, which resulted in improved performance indicators in most measured domains. It is, therefore, suggested that the observed continuous improvement in the selected measures presented here is not exclusively the result of "natural improvement".
MHS' performance monitoring system was a necessary but insufficient element for the explanation of long-term care improvement. Regional "Quality Teams", comprised of physicians, nurses and other health professionals in managerial positions, were set up and trained to guide analysis of quality gaps and implementation of effective interventions. Resources were allocated to intervene in units which had exhibited wide gaps between actual performance and desired targets. Considerable effort has been invested in empowering patients throughout programs to increase treatment adherence, among other steps taken.
Our study also demonstrated association between significant reduction in mortality and improved glycemic and cholesterol control. Data from the United Kingdom has shown that the mortality risk among patients with Type 2 Diabetes is 1.6 times higher than that of the general population. Landman and colleagues have reported that patients with diabetes evidencing poor glycemic control (HbA1C > 9%) exhibited a hazard ratio of 2.21 for total mortality, compared with a hazard ratio of 1.0 among the control group with normal glycemic control levels (HbA1C < 6.5%). This suggests that in order to increase life expectancy, interventions should focus mainly on patients evidencing poor control. The literature also shows that correction of dyslipidemia (such as control of LDL-C) in patients with diabetes promotes reduction of macro-vascular disease, which contributes to cardio-vascular complications and shortened life span.
The models for ED visits (as a dependent variable) calculated in this study were too weak to produce significant statistical results. HbA1C performance (as an independent variable) was not found to be associated with either hospitalization days or death. It seems logical that this process variable is insufficiently powerful to explain these two outcomes. Additionally, HbA1C is measured as performed at least once a year; the findings indicate that testing only once a year is insufficient for disease control and achievement of desired outcomes.
Providing appropriate care for patients with diabetes, especially those exhibiting complications resulting from poor disease control, demands considerable resources. Improving the quality of care to patients with diabetes and achieving better health outcomes is also costly. Moreover, health care systems around the world are facing pressure to constrain costs, given the rise in medical sophisticated technologies and the aging of the population, among other reasons. Those trends prompt health care organizations' decision makers to expect a "business case" for quality improvement, meaning that these investments would have a "return on investment" (ROI) within a reasonable period of time. The annual average result of 2% reduction in hospitalization days through the reduction in poor glycemic control is a preliminary pivotal evidence for such a business case. The results comply with the results of several other studies, whose authors conclude that sustained reduction in HbA1C levels among patients with diabetes is associated with significant cost savings within 1–2 years of level reduction. MHS, as all other Israeli HMOs, is characterized by a very low patient turn over, meaning that only a small portion of the population leaves its HMO during its life time. This fact urges the Israeli HMOs to invest in quality improvement, knowing that they can return their investments in the long term.
Blumenthal and colleague argue that a leading obstacle to achieving quality in health care is the absence of a "business case" for quality. Healthcare system infrastructure is frequently accused of being inadequate to support such thinking. Furthermore, one of the root causes mentioned is the primitive quality measurement stage of science; if most healthcare providers are unable to estimate the total cost of investing in quality, how can one expect them to calculate the savings produced by their investment in interventions? Also, interventions to improve diabetes care produce return on investment only in the medium- to long-term (delayed savings). Therefore, healthcare organizations with a high turnover may not be able to achieve this return. In times of austerity, the majority of budget cuts take place in the healthcare sector, which adds pressure on organizations to economically "justify" quality improvement investments.
Study's Strengths and Limitations. The strength of this study lies in the fact that data were analyzed for a very large study population (N = 96,553) comprising all adult members of a major health plan, registered in the Diabetes Register, which increases the statistical validity of the findings. In addition, a robust statistical techniques were employed to support the hypothesized results.
Yet, the study has considerable limitations, which should be overcome in future studies: (1) The effect of only three diabetes performance measures was studied. Appropriate follow up with process measures such as eye and foot examinations or medical attention for nephropathy as well as intermediate blood pressure control were not included in the analysis because the measure's definition changed during the study period or the measure was not documented throughout the study period; (2) our information system, although fully computerized, was not designed to collect accurate and full data on the costs of quality improvement interventions or medical expenditures directly related to diabetes care; hence, the "investment" side of the "business case" argument was not thoroughly looked at. The data cannot be appropriately used to substantiate returns on investments in quality. Furthermore, the cost of hospitalization is subject to local agreements and contracts between HMOs and hospitals, and is not publicly transparent. The estimated cost saving is, therefore, based on the price list published by Israel's Ministry of Health, which may not fully reflect actual prices. 3). This study investigated the entire MHS' population; thus, a control group was not available, as all MHS members were under the quality improvement scheme. A further investigation with a case control design is recommended.
Discussion
Key Findings
Our study has demonstrated that over a 6-year-period, improvement in glycemic and cholesterol control was associated with significant decreases in hospitalization days, mortality and direct medical costs. ED visits failed to demonstrate a similar association.
Results in Context. These results may be explained by the fact that improved glycemic and cholesterol control were important elements in the organization's efforts to improve quality of care, through multidisciplinary health promotion programs created within the framework of a performance monitoring system that implemented the organization's "call for action".
In the early 1990s, the Institute of Medicine (IOM) defined quality as the "degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge". Quality of care can then be evaluated on the basis of structure, process, or outcome. When used appropriately, both process and outcome measures can provide valid information about quality objectives.
The decision to choose hospitalizations, ED visits and mortality as health outcomes (the dependent variables) was based on data availability. Data on diabetes complications and long-term trends in diabetic patient satisfaction remain unavailable. It has been long claimed that the use of outcome measures is insufficient for quality of care assessment since most differences between patients receiving the same treatment result from factors associated with patients' characteristics – which are outside the control of health care providers. Following our experience, it is believed that beyond the differences in personal characteristics, differences in care provided to MHS members tend to be marginal. However, differences in personal characteristics (gender, age and socio-economic rank) were adjusted in all the statistical models. It is suggested that further research will be conducted with a control for patient characteristics such as self-efficacy, self-capability to manage the disease, compliance with medical regiment, and so forth.
Data obtained from the Israeli National Program for Quality Indicators (NPQI) in Community Healthcare indicated a 0.25% annual increase in diabetes prevalence from 2004 to 2010. National census data indicated that diabetes-related mortality rates, adjusted for ethnicity and gender, have decreased by 24% from 2004 to 2010, most probably due to better control of intermediate outcome measures and better care for diabetes complications. Although data on diabetes-related hospitalizations are not available, crude national figures indicate a steady increase in hospitalization days between 2000 and 2010.
It is worth mentioning that the NPQI, instituted in 2004, contributed to each of the four Israeli health plans, including MHS, for development of quality improving infrastructures, which resulted in improved performance indicators in most measured domains. It is, therefore, suggested that the observed continuous improvement in the selected measures presented here is not exclusively the result of "natural improvement".
MHS' performance monitoring system was a necessary but insufficient element for the explanation of long-term care improvement. Regional "Quality Teams", comprised of physicians, nurses and other health professionals in managerial positions, were set up and trained to guide analysis of quality gaps and implementation of effective interventions. Resources were allocated to intervene in units which had exhibited wide gaps between actual performance and desired targets. Considerable effort has been invested in empowering patients throughout programs to increase treatment adherence, among other steps taken.
Our study also demonstrated association between significant reduction in mortality and improved glycemic and cholesterol control. Data from the United Kingdom has shown that the mortality risk among patients with Type 2 Diabetes is 1.6 times higher than that of the general population. Landman and colleagues have reported that patients with diabetes evidencing poor glycemic control (HbA1C > 9%) exhibited a hazard ratio of 2.21 for total mortality, compared with a hazard ratio of 1.0 among the control group with normal glycemic control levels (HbA1C < 6.5%). This suggests that in order to increase life expectancy, interventions should focus mainly on patients evidencing poor control. The literature also shows that correction of dyslipidemia (such as control of LDL-C) in patients with diabetes promotes reduction of macro-vascular disease, which contributes to cardio-vascular complications and shortened life span.
The models for ED visits (as a dependent variable) calculated in this study were too weak to produce significant statistical results. HbA1C performance (as an independent variable) was not found to be associated with either hospitalization days or death. It seems logical that this process variable is insufficiently powerful to explain these two outcomes. Additionally, HbA1C is measured as performed at least once a year; the findings indicate that testing only once a year is insufficient for disease control and achievement of desired outcomes.
Providing appropriate care for patients with diabetes, especially those exhibiting complications resulting from poor disease control, demands considerable resources. Improving the quality of care to patients with diabetes and achieving better health outcomes is also costly. Moreover, health care systems around the world are facing pressure to constrain costs, given the rise in medical sophisticated technologies and the aging of the population, among other reasons. Those trends prompt health care organizations' decision makers to expect a "business case" for quality improvement, meaning that these investments would have a "return on investment" (ROI) within a reasonable period of time. The annual average result of 2% reduction in hospitalization days through the reduction in poor glycemic control is a preliminary pivotal evidence for such a business case. The results comply with the results of several other studies, whose authors conclude that sustained reduction in HbA1C levels among patients with diabetes is associated with significant cost savings within 1–2 years of level reduction. MHS, as all other Israeli HMOs, is characterized by a very low patient turn over, meaning that only a small portion of the population leaves its HMO during its life time. This fact urges the Israeli HMOs to invest in quality improvement, knowing that they can return their investments in the long term.
Blumenthal and colleague argue that a leading obstacle to achieving quality in health care is the absence of a "business case" for quality. Healthcare system infrastructure is frequently accused of being inadequate to support such thinking. Furthermore, one of the root causes mentioned is the primitive quality measurement stage of science; if most healthcare providers are unable to estimate the total cost of investing in quality, how can one expect them to calculate the savings produced by their investment in interventions? Also, interventions to improve diabetes care produce return on investment only in the medium- to long-term (delayed savings). Therefore, healthcare organizations with a high turnover may not be able to achieve this return. In times of austerity, the majority of budget cuts take place in the healthcare sector, which adds pressure on organizations to economically "justify" quality improvement investments.
Study's Strengths and Limitations. The strength of this study lies in the fact that data were analyzed for a very large study population (N = 96,553) comprising all adult members of a major health plan, registered in the Diabetes Register, which increases the statistical validity of the findings. In addition, a robust statistical techniques were employed to support the hypothesized results.
Yet, the study has considerable limitations, which should be overcome in future studies: (1) The effect of only three diabetes performance measures was studied. Appropriate follow up with process measures such as eye and foot examinations or medical attention for nephropathy as well as intermediate blood pressure control were not included in the analysis because the measure's definition changed during the study period or the measure was not documented throughout the study period; (2) our information system, although fully computerized, was not designed to collect accurate and full data on the costs of quality improvement interventions or medical expenditures directly related to diabetes care; hence, the "investment" side of the "business case" argument was not thoroughly looked at. The data cannot be appropriately used to substantiate returns on investments in quality. Furthermore, the cost of hospitalization is subject to local agreements and contracts between HMOs and hospitals, and is not publicly transparent. The estimated cost saving is, therefore, based on the price list published by Israel's Ministry of Health, which may not fully reflect actual prices. 3). This study investigated the entire MHS' population; thus, a control group was not available, as all MHS members were under the quality improvement scheme. A further investigation with a case control design is recommended.
Source...