Managed care system, public accountability, cost containment, pay-for-performance and ranking culture demand Quality of care to be monitored through appropriate instruments.

Outcome endpoints (i.e. morbidity and mortality) are still the most widely used quality indicators in thoracic surgery.

Outcomes however should be reported in the most correct way to prevent risk-averse behaviours and misleading information.

They need risk-adjustment, as different case-mixes at different institutions may influence outcome and those units operating on older and sicker patients would be penalized without an appropriate risk-adjustment. Therefore, risk-modelling must become the logical and necessary approach for provider profiling and comparative audit.

 

The most important tool of any quality assessment endeavour is a database that is made up of a representative sample of the study group of interest.

The gold standard for data should be a specialty-specific, procedure-specific, prospectively maintained, periodically audited, electronic database that contain, at the minimum, a core set of variables that has been demonstrated to be associated with outcome.

 

Although in many settings administrative databases are used to evaluate performance since they are readily available and relatively inexpensive, these datasets have many disadvantages that contraindicate their use for clinical  audit.

The most important problems with administrative data are the followings: as they are mostly collected for billing purposes, critical variables may be unavailable, there may be a difficult differentiation of comorbidities from complications, they may exclude important variables that are not billable diagnoses, limit the number of secondary diagnoses and generally have poor flexibility to properly classify certain comorbidities.

For all these reasons, claims data should be avoided whenever possible for clinical audit purposes.

 

 

The practical steps that should be planned and possibly recorded to construct a solid clinical database are a clear definition of the data sources and the creation of a list of variables (and their definitions) that will constitute the database. These steps will permit that 1) the database can be used even by subjects that did not participate to its construction, 2) the database can be audited by external data managers to assess quality of data, 3) changes in data collection or variables recording may be adequately planned.

The importance of the source and the quality of data cannot be overemphasized enough.

 

Most of the data that are of clinical interest derive from clinical records or other attached documents, such as laboratory exams or PFTs. One of the most critical aspects of the database construction is the extraction of the data from the medical record to the database. Wherever possible, data should be entered in real time, at the point of capture; to this end a networked database should be accessible in the operating theatre, the ward, the clinic and the multidisciplinary team meeting room. 

When possible this data should be used to generate documents such as operation notes, MDT report, correspondence, so that data capture becomes integral to routine patient care.

 

The person in charge of capturing or transferring data into a database should be properly qualified and adequately trained.

A Clinical Audit Lead should be selected within each unit who will be responsible for the accuracy and quality of data collection.

The data should be periodically checked for discrepancies, inconsistencies, missing values, in order to ensure a high quality database. In fact no model or predictive equation can be better than the data upon which it is based. If any underperformance in data collection would be detected this should be reported  to all persons involved in the process of data recording with the final objective of continuously improving the quality of the database.

 

Collecting data, maintaining an updated dataset, periodically assessing the quality of data may involve substantial costs. Furthermore, the logistical costs attributable to database management add to the costs of analysing data, generating and reporting results, and implementing quality measures that are suggested by data analysis.

Nevertheless, the final objective of any data collection is improving quality of care. Even if start-up costs may be daunting, ultimately improved quality of care will be cost-efficient, since the least expensive means to accomplish a task (health care delivery) is the means that employs the highest quality in the process.

Hospital administrators have to appreciate the economic importance of data collection; cost savings as a result of improved quality of care can be used to offset costs of gathering data and implementing clinical databases.

 

The European Society of Thoracic Surgeons recently appointed a Database Committee responsible to develop and maintain an online clinical Database with the aim to collect clinical data from thoracic surgery units across Europe.

A first version was launched in 2001 and collected data on all thoracic procedures until December 2003. 27 units from 14 Countries across Europe contributed consistently to the Database.

 

A first analysis on lung resection was published in 2005 and reported on the methods of development of  a in-hospital mortality risk-adjusted model for lung resection. Two models were substantially developed and validated. One incorporating subjective measures and called ESSS (ASA and Dyspna score) and the other one built with objective measures such as age and ppoFEV1 and called ESOS.

These models were derived from 1694 lung resections for lung tumors and validated on another 1126 patients showing satisfactory calibration.

The equation for the ESOS.1 model is the following: ln R/1-R= -5.8858 + 0.0501Xage - 0.0218XppoFEV1

 

The ESTS Database version 1 had several limitations and potential area for improvement:

•         Low rate of accrual

•         Only 27 units (of 120 who enrolled) contributed at least 95% complete data

•         Scarce rapresentativeness of the European thoracic surgery activity

•         ppoFEV1 and ppoDLCO calculation not standardized

•         ppoDLCO reported only in 25% of patients

•         Potential predictors missing (i.e. cardiac status)

•         30 days mortality, incomplete data, not used for modelling

•         Morbidity not risk-modelled

•         Statistics  to be refined (bootstrap, hierarchical model)

 

 

 

ESOS was recently used to assess the performance of 3 different European thoracic surgery units.

This study was performed on behalf of the ESTS Audit and Clinical Excellence ad hoc Committee,  and included 695 patients submitted to major lung resections from 2004 through 2006 at 3 centers.

Data were prospectively collected at each unit in clinical, electronic and periodically audited datasets.

Variables and endpoints of interest were selected and merged in a merged centralized dataset.

A preliminary scrutiny of the quality and consistency of variables across the 3 units was a priority of this project.

ESOS.01 was applied to estimate the in-hospital mortality of each patient

Finally, predicted and observed mortality rates were compared within each unit without showing any statistically significant differences.

 

The ESTS Audit and Clinical Excellence Committee was able to draw several Quality management inferences from this study:

•         The quality of data collection is essential for every quality management activity

 

•          Mortality may be insensitive to quality, require long term evaluations and do not provide immediately actionable quality information

 

•          Multiple risk-adjusted outcome indicators should be used for a more reliable and comprehensive provider profiling as they may reflect different aspects of quality

(cardiopulmonary morbidity, technical complications, emergency ICU admission, residual function, quality of life, etc)

 

Furthermore, future directions were set:

• Refinement of data collection to improve their quality and cost-effectiveness

 

• Development of process-based indicators of quality

 

• Use of breakthrough reliability statistics (bootstrap) for cross-validation of models

(Blackstone JTCVS 2001; Brunelli & Rocco JTCVS 2006)

 

• Research to identify the optimum time window for estimating and reporting performance

 

• Research of mechanisms for dealing with underperformance (peer-based confidential TQM projects)

 

In July 2007, the ESTS Database Committee launched the second version of the online European Thoracic Database. The Database is linkable from the ESTS website at the following address (https://www.thoracicdata.org).

A form must be filled to obtain a userID and password to login and join the database project.

The database may be used as an internal data capturing tool and for contributing the Lung Resection Risk Model Project, which aims at improving the previous risk model for mortality by adding critical variables such as DLCO and cardiac co-morbidity.

 

 

The European Database allows the inclusion of process measures of performance.

Until recently, quality improvement initiatives and health care reform were focused on the short-term outcomes (i.e., operative mortality and morbidity).

However, it is now recognized that such clinical outcomes are only one measure of overall health care quality. More comprehensive quality indicators are desirable, including intraoperative processes of care as well as perioperative measures that impact hospital outcomes, secondary prevention, and long-term health. 

 

Indicators of Quality can be related to structures, process or outcome of health care.
“Structure” denotes the attributes of the settings in which care occurs (structural organization, human and material resources).
“Processes” measure the activities and tasks in patient episodes of care. From a performance management perspective, the key issue is that a desirable process should be unambiguously associated with improved patients health outcomes. Monitoring the process can then be a substitute for measuring the outcome.
“Outcome” measures attempt to describe the effects of care on the health status of patients and populations.

 

In the future, process indicators of performance will be incorporated in the European database with the aim to provide more comprehensive instruments of evaluation of performance
The ESTS Clinical Guidelines ad hoc Committee has been appointed and will be responsible to develop evidence-based guidelines that can be potentially used as process measures

 

Quality is an abstract construct that cannot be directly measured. In the nomenclature of modern measurement theory, it is characterized by one or more latent (unobserved) variables or traits.
To quantify abstract constructs like quality, intelligence, or musical ability, we typically rely upon some combination or composite of measurable surrogates that are thought to be associated with or contribute to that underlying trait

 

Ongoing ESTS initiatives are aimed at developing Composite Performance Scores by using data collected in the European database and units will be rated according to their scores with the intent to create a platform for improving the quality of care.

 

Future developments and activities related to the ESTS Database will include:

 

•         European Database adopted by National Societies as National registry for credentialing and regulatory purposes

•         European Database used as sovranational credentialing instrument

•         European Database used to develop and test new standardized pathways of care

•         European Database as an international common platform for quality benchmarking

•         European Database linked to the ESTS Directory of Thoracic Surgery with updated data on European thoracic surgery activity . (http://www.thoracicdirectory.org/)

 

 

 

 

There are no comments to this entry.