My Pet Peeves: Generalizing Data to the Wrong Group

[1] arguing that having a homebirth is as risky as letting your child ride in a car without a seatbelt.  The authors are an Australian OB and an ethics professor in the UK.  The article is really a long-winded way to say that homebirth should not be banned, but rather severely discouraged because of the potential long-term effects and, what they conclude, as higher risk of death for infant and mother.  As ethicists, they write this for all countries.  All doctors should promote hospital births.

The question becomes, what evidence did they use to make this decision?  Well, they very rightfully note that there is virtually no research on long-term disabilities related to location of planned birth.  This is a huge flaw in the system and one every country should look into.  However, I say “virtually” because all we have to go on is one large-scale study out of the UK[2] which did examine long-term outcomes.  Now, because the base rate is so low, they lumped all long-term outcomes together because they didn’t have the power to detect any one individually.  This is a flaw because it means less severe outcomes are mixed with more severe, but this is a regular problem when baserates are low.  What did they find?  No difference in long-term outcomes for planned home versus planned hospital births for multiparous women.  In the UK.  This is obviously limited research and not something we can take to heart as being definitive, but it suggests that in a system such as the UK, planned home birth may not have any increased risk for later outcomes.  The authors, however, argue the potentials of what could go wrong because of delayed transport without any data to back up that the current guidelines about transfer times and homebirth in many countries are flawed.

Now what of death?  Well, these authors focus exclusively on US data[3] (and a supplement that I have to take their word for because apparently it’s impossible to find even online, though the same authors have written up US data suggesting a higher risk to homebirth[4]) to suggest that homebirth is inherently less safe (and use the infamous Wax et al. study[5] which is considered a gong show at best from a methodological and statistical point of view).  They don’t even include any references to the large-scale studies in Canada[6] and the Netherlands[7] that has found no increased risk for homebirth versus hospital birth and one US study looking at an integrated system (like those in Canada and the Netherlands) found the same outcome[8] (nudge, nudge, USA).  The one study they included to try and argue that travel time matters even in highly integrated areas was based in the Netherlands and looked at travel time for all births (not home versus hospital) and was based on travel time of 20 minutes or less versus more than 20 minutes[9].  Travel time greater than 20 minutes was associated with greater death, regardless of planned location as this included just traveling to the hospital at the onset of labour.  Given this is the time frame suggested in Canada too for a homebirth, it seems that it’s based on some evidence.

 

This is where my pet peeve comes in: When you’re looking at either mortality or morbidity from birth, you’re looking at an event that will be categorized by the health care system in which one lives.  How a health care event unfolds is therefore going to be highly dependent upon the system in which it resides.  As ethicists, I would sincerely hope that these individuals were aware of how ludicrous it is to take US data and use it to make recommendations for the rest of the world, especially when we have large-scale data from other countries that counter US data because of different health care systems.  Furthermore, when you use data from one country with a health care system different from the rest and you try to demonize homebirth and the women who choose it (borderline suggesting it is immoral and should be illegal), you have a lot of explaining to do.

The authors state, “What we have argued is that a neglected ethically relevant outcome is serious avoidable, foreseeable disability to a future child.”  Except later when they have to admit, “We have argued that in most situations the risk of disability may be elevated and unjustified.” (Emphasis mine.)  In short, they don’t have any evidence this is the case, but assume it might be.  What evidence they have suggests that there may be concerns dependent upon how far people live from the hospital (why many care providers in countries with an integrated health care system have distance from hospital discussions with women and their families), and they are right to suggest more research into it.  What they cannot do is make conclusions in the manner they have from the evidence they have.

They also pull out the “one child harmed is too many” while ignoring the risks that may be present in a hospital (though they do give lip service to the “possibility” that hospitals can have risks too), playing it off as if only homebirth could be associated with negative outcomes.  I actually couldn’t help but snort when I read, “[P]rofessionals should not support planned homebirth when there are safe and compassionate hospital-based alternatives.”  I don’t doubt there are safe and compassionate hospitals out there or even compassionate doctors in hospitals that generally aren’t, but when a large portion of women are looking for homebirth because their hospital experiences were the antithesis of compassionate, these ethicists need to be looking at what they are suggesting.

Women and their families deserve to be given every bit of information they can, but that information should be specific to their health care system, their hospital, and their midwife or doctor or whoever they have providing support.  Telling a woman in Canada that homebirth is riskier because it is in the USA isn’t true and certainly isn’t sound science.

 [Image Credit: Unknown]