New perspectives of assessment: good health outcomes in communities with unsatisfactory care

29 Jan

Mireia Espallargues Mireia Espallargues. Head of Healthcare Quality AQuAS

Identifying patients with good results in communities with unsatisfactory care can be the key to finding success factors.

There are several initiatives that advocate a high value care in order to contribute to greater efficiency and sustainability of health systems. This value is interpreted as obtaining better health gains relative to the costs, which translates into better use of available resources (1). The more traditional approach to detect this potential value was based on the identification of patients with poor health in order to establish subsequent remedial measures that lead to a good result.

In an article published in The New England Journal of Medicine, Sequist and Taveras (2) propose to analyze the problem from a different perspective:

  1. Identify patients who are doing good (“positive outliers“)
  2. Analyze what factors may influence their good health
  3. Disseminate the identified success factors and expand them to the rest of the system

To accomplish this, the authors propose a new approach to measure and analyze information. It’s about relating the supply system, the community and the patient in an effort integrated as a strategy with the aim of improving population health. This type of approach strengthens assessment models that AQuAS is applying in areas such as attention to chronicity and integrated health and social care in our country (3-5).

Health determinants and social determinants

According to the authors, the population’s health status is the result of a complex web of interactions between individual patient characteristics and other determinants of health dependent of the area of residence (community). The more evident is that most determinants that affect the health of people would be generated outside the health service delivery system. These factors, that some studies estimated constitute about 80%, have a profound effect on how the patients interact with the system and consequently, the quality of care they receive and their health outcomes.

Thus, the policies and the actions in areas such as housing, employment or social welfare can not be decoupled from assessments on health and from the health system. We must emphasize the need of introducing in the context of public and social policies, the concept of health as an item to be referred to in all its dimensions and evaluated in an integrated manner, together with other social determinants besides the health system (6).

The challenge of evaluating integrated care

Given this new approach it’s necessary to analyze the performance at the community level, usually focused on the area of residence. This analysis will allow provision system to obtain a better understanding of its population and identify where patients are grouped within communities, as well as what are the environmental factors that can affect health outcomes. This represents a quality leap beyond traditional analysis of each service providing system units (hospitals, primary care, care teams, etc.). Among the advantages of this new strategy we must specify the identification of patients who reside in communities where the quality of care and the results are not satisfactory, the detection of promising approaches for patients in these communities and integrating these successful strategies into care plans for the patients.

The definition of the community and the analysis of data for evaluation

To perform this type of assessment it’s necessary to first have a good working definition of “community” as well as a robust infrastructure for data analysis. The second step consists in identifying positive outliers – extreme values with good results- in communities with poor performance or with a high burden of disease – hotspot communities– more specifically, the identification of treated patients with good health outcome in an environment of unsatisfactory service provision and, in particular, those who historically had poor results and who recently improved. Once the success factors are identified, we would move on to the phase of climbing or extending the strategies in the rest of the territory.

In the precise case of the project assessment of the care for chronic diseases, the territories with programs or models of care to chronic diseases seeking greater care integration have been identified, given the large impact on morbidity and mortality and the use of resources caused by a complex chronic population. Consequently, the analysis from this territorial or community vision -taking into account the various healthcare resources and interventions in the territory- has allowed us to identify programs or models that perform better in relation to a number of health result indicators of various quality of care dimensions. This way we can select those that are “outstanding” (good outcomes for patients) and then identify good practices and success factors.

The last step is the integration of these strategies or models with the plans of patient care. Sequist and Taveras cite some other examples of community initiatives that relate to clinical practice (care or welfare services provision) as communicating vessels between the data analysis and the interventions that can be made in various areas.

This analytical approach can have several potential uses and can also be a powerful tool for addressing socioeconomic inequalities in health outcomes; as long as they focus on the differences at the context level, on patients’ membership to a community, on differences in gender, income, or education. Finally the authors mention that, as a prerequisite for a successful implementation, it’s necessary to have a well-defined operational infrastructure in which funding is aligned with the approach of linking the community and the health care.

Similarly, sustainability challenges also occur, as it may be that the service providing system “buys” this concept of factors that historically have been considered outside the area of influence or responsibility of health care. We must also ensure that resources and community interventions are safe and reliable if we want to have the support of professionals to refer patients to these resources as well as determining the most appropriate information updating intervals.

To conclude, we’re looking at an approach that instead of focusing on the “non-compliant” patients, it’s based on the observance and analysis of the best, especially in disadvantaged areas, with the purpose of applying the same keys to success to other territories.


(1) Porter ME. What is value in healthcare?. N Engl J Med. 2010;26:2477-81.

(2) Sequist TD, Taveras EM. Clinic-community linkages for high-value care. N Engl J Med. 2014;371(23):2148-50.

(3) Desenvolupament d’un marc conceptual i indicadors per avaluar l’atenció a la cronicitat. Primer informe. Barcelona: AQuAS; 2013.

(4) Consens i selecció d’indicadors per avaluar l’atenció a la cronicitat. Segon informe. Barcelona: AQuAS; 2013.

(5) Serra-Sutton V, Montané C, Pons JMV, Espallargues M. Avaluació externa de 9 models col•laboratius d’atenció social i sanitària a Catalunya. Barcelona: AQuAS; 2014 (en premsa).

(6) Determinants socials i econòmics de la salut. Efectes de la crisi econòmica en la salut de la població de Catalunya. Barcelona: Observatori del Sistema de Salut de Catalunya. AQuAS; 2014.

Medical practice: do a lot or do just what is needed?

22 Jan

Joan-PonsJoan MV Pons. Head of Evaluation AQuAS

Medicine is an imperfect science (hence it’s an art) but it’s also a craft, and it has always tended, perhaps more for the latter than the former, to want to act, to do something, even to no avail, to the extent that it often causes serious damage (bleeding) or if there is a beneficial effect, it is often not apparent until years later (Jenner’s smallpox vaccine); it acts with the aim of demonstrating that something is being done; moreover, adding successive interventions to show that everything that could be done has been done. Why this desire to act and to do more each time?

There may be several reasons. One of them, a simple one, is linked to the profession itself and to private practice (private insurance) is that; if you’re paid to do it, you end up doing more than required, especially when it comes to diagnostic tests; always with the best of intentions so as not to forget anything, collecting the maximum information with the aim of covering any remote possibility of any unlikely diagnosis. Defensive medicine emerges from here and occurs more in countries with over employed lawyers where the fear of litigation –founded on increasing casuistry- leads to an increasing number of tests. The case of Dr. Daniel Merenstein, a 3 year resident with a patient and the PSA is well known (you can read it at “Winners and Losers”, a JAMA “A piece of my mind” Section article).

It doesn’t stop being a vision that completely neglects the other side of the coin, as if the diagnostic tests were harmless, as if, leaving aside the ionizing radiation from many imaging tests, there was no risk of false positives, false negatives with all their consequences; something similar to countless laboratory tests and biomarkers within reach. There is no perfect test that discriminates with 100% accuracy, nor is human nature itself, always heterogeneous in the extreme (fortunately rare).

It is sufficiently well known that the more variables that are explored, the greater the risk of finding significant results simply by chance; hence the need to correct the level of statistical significance (the famous “p”) in studies with multiple comparisons, such as genetics where they analyse many polymorphisms. Many of these genetic studies of broad scope, generated mostly by technology and its increasingly lower cost, than for a prior hypothesis (authentic fishing expeditions), to rule out pure chance, should apply extremely low values of statistical significance. With the proliferation of imaging tests with a higher and higher resolution, we now speak of those “incidentalomas” to describe those incidental asymptomatic findings exposed while looking for unrelated things; no need to specify the ethical and practical implications that this entails. It’s the same when genetic tests are requested indiscriminately.

There is another reason that can be invoked in order to explain this increased desire to add more than to subtract. Psychologists and economists, academic areas that grow increasingly closer, talk about loss aversion in the sense that we are more affected by the loss (what we had and we no longer have or what we were doing and we are no longer doing) than by the possible gain. That’s why we find it so hard to abandon practices, many simple routines, as one engages autopilot, which do not provide us with any useful information or maybe even worse, may pose a risk and an unnecessary expense.

Excessive medicine, without critical reasoning in acting, can be neither good for patients or for the health system. Don’t forget that when we speak of a health system, we are referring to a universal insurance (for all people) and that this allows us to spread the illness and financial risks of an increasingly expensive health care system. Most of the health budget however, is in the actual hands –the real hands not figuratively speaking- of health professionals when they’re applying tests or prescribing treatments. When the interventions (preventive, diagnostic or therapeutic) for a particular individual do not add value, we are also wasting the shared and limited public resources.

Can a composite indicator summarize hospital care quality?

14 Jan

Josep Maria ArgimonJosep Maria Argimon, Director AQuAS

The “Central de Resultados” (Results Central) mission is to measure, evaluate and distribute the results achieved by different public health care system centres. The aim is to facilitate better decision-making with regards to the quality of the health care service provided to citizens. The reports feature results from patient satisfaction to the efficiency in the use of resources. The ultimate goal is to foster an internal review of the elements that influence the results, especially in organizations that repeatedly show indicators which are worse than others. To encourage this task, the institutions are nominally identified: not as a punishment but as an incentive to improvement.

Last November the “Agència de Qualitat i Avaluació Sanitària de Catalunya” (AQuAS) released the results for 2013 for hospital and primary care, social care mental health and addiction treatment. The “Central de Resultados” report for hospital care consists of 99 indicators and presents some new features over previous editions. One of the most important is the inclusion of a synthetic indicator of adequacy, in which avoidable hospitalizations for heart failure, diabetes or chronic obstructive pulmonary disease weigh heavily, but where other indicators are also given their due significance. Given its complexity, this new indicator of adequacy was the subject of a detailed explanation during the public presentation of the results and, at the same time, is one of the most featured in the media. In this case, among other centres, the Hospital de la Santa Creu and Sant Pau are examples of the hospitals which should improve this indicator, and thus the media picked on the aforementioned report.

Summarizing the quality of health care provided by a hospital based on a single indicator is very difficult. Now, since the media highlighted the results of this particular indicator, the responsibility for it shall not rest with the messenger but with the issuer, in this case, me. Surely, the communication was poor because if we continue to analyze the report to its full extent we can observe other relevant indicators. For example, the indicator of mortality at 30 days after discharge for selected causes – another novelty of the presented report – shows how a hospital that let’s say has poor results in the synthetic indicator of adequacy, can at the same time have excellent results in the indicator of mortality, such is the case of the Hospital de la Santa Creu and Sant Pau. This is an indicator that is made public in very few public health systems, and some agencies, among which American and British, have promoted its use. Similarly, in other indicators, such as satisfaction with the emergency services, the same Hospital de la Santa Creu and Sant Pau obtained the top positions, as has been well reflected in the media.

We have to keep in mind that the same indicators, despite many efforts, do present some shortcomings, starting with the absence of important outcomes such as the functional ability of patients and their quality of life related to health. But there’s something else that also escapes and that only very specific studies can analyze. I’m talking about the culture of the organization that is largely configured for all the staff that belongs to it; what values they share and the feelings they have towards it. Those organizations that – despite facing serious external problems and discouraging  states of mind, present in any collective, –  believe in the values that inspire their work;  have as their objectives continuous quality improvement; where everyone is involved regardless of age or position (from the manager to the warden), have a better communication and coordination between professional and technical groups;  know without shame or guilt how to examine the causes of their mistakes or failures;  participate and get involved in improvement proposals; all these difficult to collect and analyze features at the organizational and staff level are at the moment of truth, the ones making a difference.

The Results Central is not a perfect instrument, but provides data for evaluating public policies, helping to insure that the established objectives are met and that an efficient allocation of public resources is made. Furthermore, transparency in policy making and accountability in the management of public resources is a requirement of citizenship, a democratic and ethical obligation of governments, and a key element in the development of a country.

The Department of Health, through AQuAS, does not avoid discussing sensitive issues, such as the reports from the Central de Resultados presented last November or the report on the effects of the economic crisis on the population’s health. The press has always addressed these issues with appropriate importance, seriousness and rigor. Therefore, it’s a shame that due to an issuer’s (not a messenger’s) miscommunication they can misinterpret the results and end up detracting from the outstanding work performed by professionals in extremely difficult conditions, so that the thousands of citizens served, cared for and cured on a daily basis know they are in the best possible hands.

Expensive drugs with limited effectiveness: what are we willing to lose in order to be able to pay for them?

8 Jan

AAnna Garcia-Altésnna Garcia-Altés. Head of the Catalan Health System Observatory

Some last generation biotech drugs such as Sovaldi (Sofosbuvir) for hepatitis C, or Avastin (bevacizumab) for breast cancer are putting the health system at a crossroads. Although the price of Sovaldi is still unknown, by projecting the price obtained in other countries over the prevalence of hepatitis C in our context, we can forecast an outrageously high figure. Meanwhile, since its release in 2004, Avastin has been a top selling drug that in a few years has obtained the approval for use in the treatment of five tumour types and its use in fifty others is being investigated.

In the United States, it’s estimated that biotech drugs are responsible for 70% of the increase in drug expenditure in recent years. The turnover’s relevance is very often reflected in the newspapers’ health pages. This set also encompasses drugs for rheumatoid arthritis and multiple sclerosis.

The issue transcends the anecdote, as shown in an interesting article last month about the value of the new drugs. At reviewing effectiveness (in terms of years of life adjusted to quality- -AVAQ-) and cost-effectiveness ratio of the drugs approved by the FDA between 1999 and 2011, Peter Neumann and his co-authors found that most of the drugs provide modest benefits compared to the traditional drugs: 32% showed no additional benefit; while from those who did, a third reached figures below 0.1 Incremental AVAQ (which would be equivalent to five weeks of quality-adjusted survival), and two-thirds less than 0.3 AVAQ (15 weeks). There are also 14 last generation drugs and 2 traditional drugs that offered more than half AVAQ benefit (6 months quality adjusted survival). However, its cost-effectiveness is also much higher: 2 of them had less than € 50,000/AVAQ ratios, 3 between 50,000 € / AVAQ and 100,000 € / AVAQ, and 6 above € 250,000 / AVAQ. All this does nothing more but reawaken the debate on the threshold of cost-effectiveness value that, as a society, we’re willing to pay.

Sovaldi and Avastin cases remind us that society can not avoid taking difficult decisions between technologies that may improve health. Despite its problems, the cost-effectiveness threshold is a useful tool for making well informed decisions, and should be used with seriousness and consistency. It’s possible that, at the current price and in certain population groups, Sovaldi may fit into the traditional limits of cost / AVAQ, but its widespread use does raise some serious doubts about the likelihood of a return as well as many questions about which services should be stopped in order to finance it.