Registries don’t work for doctors

Tagged registry, SurgicalPerformance, Audit, quality; Posted 11 years ago

In order to measure and improve the quality of health care, professional societies and Colleges (on behalf of government organisations) feel a strong need to collect data. “Big data” inform about the current situation in the diagnosis and treatment of a variety of medical conditions, will identify outliers and enable those bodies to manage them.

The easiest way for those organisations to capture data is through a “registry”. The incentive for doctors is that they receive feedback on their outcomes. 

Recently the NHS made data collection mandatory for all its surgery. Not enough, the NHS also published their doctors’ outcomes under the protest of a large number of its medical staff. For the last couple of weeks, the NHS informs the public about “the best and the worst” doctors: “Name and shame”. However, some doctors made use of a small loophole in the English legal system; collection of data comes under privacy regulations and some surgeons declined their consent for their outcomes to be made public. Consistewnt with naming and shaming, these surgeons are named by the NHS and shamed for their unwillingness to have their outcomes published. The Mexican standoff continues. What is this relationship between NHS and its employees like? Would I love to work for such an employer? Does the NHS treat their people like their greatest assets who they endeavor to support or like enimies within?

While no one argues that accountability of surgeons is critical to good outcomes (What you don't measure you cannot improve"), the use of the stick and punishment does not advance the cause. Both, doctors and patients are interested in best surgical quality. However, I do question the way it’s currently done overseas and even in Australia.

I give you an example from a Danish registry. The Danish Lymphoma Databse was established and collected "big" data. Their mangers were proud of their achievements until someone mentioned that maybe not all patients get entered into the database. Even worse, a suspicion kept coming back that selected patients do not make it into the database. Those patients were older and with poor prognostic factors; in brief, these were patients who were expected not to do well, and not quite fitr to tolerate aggressive treatment.

Why would a doctor who treats patients with lymphoma, and who is supposed to enter all patients (unselected) into a database, withhold data entry of patients with potentially poor outcomes?

The Australian Joint Repalcement Registry is another example. I had a chat with an orthopaedic surgeon in the tearoom of my operating theatre. The joint registry is a large database that records all (???) cases of knee and hip replacement. It is financed by surgical device companies that sell joints to hospitals and surgeons and is supposed to keep surgeons on their toes.  Informed consent is required and has to be obtained by patients.

I asked a local orthopaedic colleague and confronted him with gossip that suggests that some orthopaedic surgeons would not enter certain, specific patients. He had no issue confirming that patients who developed a surgical complication postoperatively are more likely not to be entered into the database than patients who have an excellent recovery from surgery.

Why would an orthopod who does joint replacements and who is supposed to enter all patients (unselected) into a database, withhold data entry of patients with potentially poor outcomes?

They don’t enter into the database because they are scared. Rightly scared. In the Danish example one of 9 units recorded poor outcomes. Statistical adjustments for risk factors neutralised that concern and that particular unit did well. It had a very normal and average complication rate. By contrast, another unit who appeared to have a normal complication rate, had significantly poorer outcomes when adjusted for risk factors.

Doctors are concerned about knee jerk reactions from media savy burocrats; doctors are concerned about the possibility that an incompetent health administrator or an incompetent College committee chairperson recommends that a surgeon must be audited and while s/he gets audited needs to pause all surgery (which can easily take 6 months). These things actually did happen in the past and destroyed careers.

Registries cost a lot of money and don’t work because data collection needs to include the need of the surgeons. Governments buddy up with professional societies and Colleges who (in the name of greater good) weed out the “bad guys”. This concept of name and shame is flawed and only works for the government but not for everyone.

What should happen is that doctors look for ways to be in charge themselves about their own outcomes. Should my outcomes be dropping, I should indeed be aware of it. At the moment I would not even know about my outcomes if I would not audit my practice meticulously myself.

More than 120 doctors participate in SurgicalPerformance right now and have entered more than 5000 cases in gynaecology, obstetrics and colposcopy. The information captured in SurgicalPerformance is for the doctors’ eyes only. SurgicalPerformance is a platform where sensitive information is collected, processed and returned to its users in strict confidence. The information is used as a learning opportunity for users and no one else than the user can identify her/his outcomes.  No user will ever be identified or reported. In SurgicalPerformance our primary loyalty is with doctors. Relevant metrics and the anonymous nature of the information are critical to forming a constructive learning environment.

In the absence of alternatives, registries are a cheap trick to coerce doctors into data collection. Registries work for those who run them but not for doctors. They are not learning but monitoring tools. We need tools that put us in charge and actually help us become even better doctors.