Dying in hospital

The neurologist Oliver Sacks published an article in the New York Times "My own life" in which he announced that he has just been diagnosed with a liver metastasis. "I find myself facing death. It’s up to me to choose how I will live the remaining months of my life. I have to live as good, as deeply and as productively as possible." Death is an inscrutable fact that each person has to go through in their own way, so far it’s all normal, and Sacks dealt with it calmly, taking the reins of the time left. But how about the health system? When someone calls the ambulance because they have a sudden chest pain, the entire health organization is tense, protocols are activated and everyone knows what to do. The health system has full control of that process so all that’s left to do for the patient is to remain confident. But on the other hand, when what’s left to be done depends on how the patient sees life and death, as is the case with Sacks, the system no longer feels so confident and can act inappropriately, even disproportionately.

As a result, the system does not quite generate a rigorous reflection on how to act when it comes to the end of life process, and therefore does not know how to interpret mortality indicators. Examples are: mortality rates of hospitals, one of the ratios which is usually listed in the categories of quality or even in those of effectiveness. It is a robust indicator and a relatively simple calculation. For example, standards warn that for each 25 people admitted into the hospital, one dies. This is a rough fact, but despite that it can be adjusted for complexity, or even to adhere to the conditions of increased risk, in the end the problem is interpretation, because when a hospital has a high mortality rate, there are always structural explanations: "in this country we’re lacking in the long-term resources" or "we have a very aging population" and the debate ends quickly.

Dr. Roger Stedman, an intensivist, now a medical director at an English teaching hospital has typified in his blog three different kinds of deaths in hospitals:

The inevitable and expected deaths

When the cases of people who have died in hospitals due to planned periods of end of life are discussed, questions that should be asked are not the traditional quality in the sense of analyzing what has gone wrong, but: Could you have avoided that person dying in hospital? And now after the fact has occurred, the second question should be: Has he/she had a dignified death? And the third: Are there protocols on how to address the end of life processes outside the hospital units? Therefore, if we had this indicator (we don’t), a low rate could be related to a good level of organizational quality among community services, primary care, social health and hospital, and if the rate was high, the interpretation should in reverse.

Unavoidable but unexpected deaths

There are processes such as myocardial infarction, or stroke, or femur fractures, which are known to be of high risk to the lives of people who are hospitalized with these tags. For example, the statistics say that for every 100 people admitted in hospital with myocardial infarction, 7 will die, for stroke the ratio is 14, for femur fracture is 5. But we know that in a well-run hospital these figures may be significantly smaller. For this reason, before a hospital death for some of these causes, or similar, the first question to ask is: Have we done everything we could? And successive issues have to do with the internal organizational quality, such as: Do we have CPR protocols (cardiopulmonary resuscitation)? Are the internal transfers qualitative? Have critical services functioned properly? And so on. In this area, there is indeed a question of internal organizational quality and this indicator should be separate from above in order to be able to analyze it properly.

Avoidable and unexpected deaths

In hospitals, sometimes things go wrong due to lapse, incompetence, inexperience, negligence or organizational errors, many of them due to communication failures between services or between shifts. This is the home ground of clinical safety programs and error analysis, rather than indicators, and the only question we should ask ourselves before each case should be: What could have been avoided? Each failure should be a lesson in order to avoid mistakes becoming established and perpetuated into the functioning of the institution. "To err is human" says the tagline, but there are hospitals with fewer errors than others, and this is due to an internal work based on an attitude of continuous improvement. Stedman offers a very timely advice: because 70% of deaths in this section are ultimately due to sepsis, it would be very effective if the hospitals would have an effective protocol to act the most preventively way possible at the first signs of sepsis.

My proposal, based on Stedman’s classification  is that we should start cataloguing hospital deaths according to these three categories, and then everything would make sense. We would know that the first group would evaluate the position of the hospital in front of the end of life processes; in the second group, internal organizational quality; and in the third, professional and organizational failures. Let's do it.

Jordi Varela

Editor