Consent and Treatment Cascades: The case of giving birth

Apologies for my long absence – moving to Australia turned out to take up quite a bit of time, and make it hard for me to access and write on the blog… Still now I am here and in far more practical role than my previous appointment (I’m now in a medical school rather than a philosophy department) I thought I might make a relatively practical post…

A medical treatment cascade is where one intervention leads to another and another and so on. Theses can occur in several contexts, for example with my first wife Kerry who had Cystic Fibrosis, she took morphine to aid her breathing. This had the side effect of causing her severe nausea for which she took another drug which itself had side effects of dizziness and lightheadedness for which she took another drug to counter act (ironically this other drug actually blocked the receptors that morphine works on rendering it useless as a treatment in the process, something we discovered only after a review of all her medication a year later). Likewise it can be the case that a routine scan picks up on something that wasn’t being searched for, and that this leads to an operation which itself has complications to remove something that may well be benign.

However perhaps the most well known treatment cascade is in regards to the delivery of children in hospital, which is commonly referred to by the natural birth movement as the cascade of interventions. (While I don’t intend to endorse everything the natural birth movement suggests – particular the more radical fringe who are in favour of unattended birthing – nonetheless, research seems to support the existence of this phenomena). The cascade here happens in part because of the psycho-social aspects of child birth (in particular that feeling comfortable, confident and relaxed contributes to smoother births) and in part because some interventions improve safety in some regards but simultaneously are likely to hinder progression of labour. The classic cascade of interventions in child birth goes like this.

In the early stages of labour you arrive at the hospital wanting a relatively natural birth in a safe environment. You are offered continuous fetal heart monitoring to determine whether the foetus is distressed at any time and thus whether the process needs to be hurried along, which you consent to because of course the safety of your child is paramount to you. This either involves having a belt strapped around your waist or involves having an electrode “attached” to the head of the infant – in either case this significantly limits the ability of the woman to move around, change position and be comfortable, making it likely to delay the progress of labour. So given that labour doesn’t appear to be progressing six hours later you are offered pictocin to induce labour and hurry things along. This works however one of the side effects of pictocin is significantly increased levels of pain, and while you had intended to not have pain killers, you reluctantly consent to an epidural after another 3 hours because it is simply too much for you. This has the effect of removing the pain, but also your ability to feel what is happening “down there”, so progress again stalls. Later exhausted after 36 hours of labour you consent to a caesarian section – with typically much higher risk of death and worse outcomes for both the mother and the child.

 

There isn’t much work on these treatment cascades in the ethics literature, but they seem to me to be of moral concern for three reasons: 1. they predictably lead to unwanted/worse outcomes for patients, 2. resource allocation, and 3. Informed consent.

 

I’m going to focus on informed consent but I will briefly describe the ethical issues in regards to the first two concerns. So firstly we might be concerned about reducing or minimising treatment cascades for patients because they are typically associated with worse or unintended outcomes. Now of course not all cascades are predictable, but where they are this gives us a reason to at least attempt to minimise them. Secondly treatment cascades are also problematic in terms of resource allocation, because typically they lead to sub-optimal outcomes (in the case of pregnancy C-section rates are far higher than they should be in most Western countries and this predictably leads to worse outcomes) and because each step along the cascade represents an investment of health care resources (another reason why we shouldn’t think about resource allocation as a series of one off cases) leading in the worst case (for example as happened with Kerry) a significant waste of resources where money is being expended for literally no patient benefit, since the various treatments simply cancel each other out.

 

The reason that treatment cascades are worrying in regards to informed consent is that while you consent to the first treatment, you do not consent (at that stage) to the further treatments it leads to. This is problematic because typically cascades present as emergency or urgent situations and hence the quality of consent can be quite low, with the patient feeling rail roaded into a particular decisions because the treatment is needed. This is particularly problematic when the risks of each intervention are progressively higher and more serious, since the initial decision seems so low risk that it may have been made relatively lightly without reflection on the possible consequences, but ultimately leads to a high risk decision.

 

So what can be done?

First and most obviously we can resist ending up in cascades in the first place – this is particularly the case when the initial intervention which leads to the cascade is of minimal medical value – in the case of routine continuous fetal monitoring for example outcomes are no better than that of traditional methods of monitoring the child’s heart rate intermittently – via either listening with a pinard stethoscope or via doppler – while they may improve outcomes when there is reason to already be concerned, as a routine measure they decrease rather than improve outcomes. The main benefits are managerial – a smaller team can manage more women using continuous fetal monitoring because they don’t need to be present regularly to check heart rates – but that doesn’t seem an overwhelming reason for recommending what is arguably a worse standard of care. But of course not all interventions that lead to cascades are of this nature. In these cases I’m inclined to think it would be better practice if the risk of the cascade was flagged up front for the patient, so that the initial consent contains some awareness of the possible risks down the track. While this will still be non-ideal in terms of informed consent, it is at least better.