A short and anonymous report just appeared on the World Health Organization (WHO) website. It is titled “Summary of prevalence of reported congenital birth defects in 18 selected districts in Iraq.” Previously, this report was referred to on the WHO website as a “joint study” with the Iraqi Ministry of Health (MoH) which began in May-June 2012. It was to examine the prevalence of congenital birth defects in a number of geographically dispersed areas of Iraq which were exposed to bombardment or heavy fighting, or were unexposed.
This joint investigation was initiated following widespread public alarm over unusual increases in poor reproductive and birth outcomes in Iraq after the US-led invasion. Across Iraq, increasing numbers of birth defects appear to be surfacing, including in Mosul, Al-Ramadi, Najaf, Fallujah, Basra, Hawijah, and Baghdad. In some provinces, cancers also are rising. Sterility, repeated miscarriages, stillbirths and severe birth defects – some not found in any medical books – are reported widely.
This explains why many public health scientists awaited the release of the WHO/MoH report on birth defects in Iraq.
In a BBC documentary which aired in March 2013, “Born under a bad sign”, a senior official of the MoH speaking on camera, said, “All studies done by the Ministry of Health prove with damning evidence that there has been a rise in birth defects and cancers” in Iraq.
During the same interview, two other MoH researchers confirmed that the situation with cancers and birth defects constitute a “big crisis” for the “next generation” of Iraqi children. In fact one researcher, pointing to a colour bar chart, said that cancers and birth defects are increasing simultaneously in three areas. As she pointed to the peaks in the figure, she named these areas “Nineveh, Anbar and Najaf”.
In the September 2013 report, Mosul (the capital of Nineveh Province) appears at the top of the list. Nineveh has the highest rates of spontaneous abortions (miscarriages), stillbirths, and congenital birth defects in the country.
It is shocking to see this report declare “no clear evidence” for any abnormality in rates of “spontaneous abortions”, “stillbirths”, or “congenital birth defects” anywhere in Iraq.
Even though data analysis is prone to variations in output, which can lead to potential changes in conclusions, for a change of this magnitude – from “damning evidence” to “no clear evidence” – extensive data manipulation must have taken place.
What happened to the data between March and September? Even though data analysis is prone to variations in output, which can lead to potential changes in conclusions, for a change of this magnitude – from “damning evidence” to “no clear evidence” – extensive data manipulation must have taken place. How and why the data was manipulated to render such drastically different results remains unknown to us.
The ultimate question to be answered is, how can this analysis, these results, and these conclusions be believed? In March, the same MoH reported “damning evidence” of a rise in Iraq birth defects. Now in September, this new report must be viewed with extreme caution if not with suspicion and disbelief.
Another unusual and outrageous feature of this report is its anonymity. No author(s) are listed or identified. An anonymous report is rarely seen in epidemiological reporting given the multiple questions that often arise when interested readers examine complicated study designs, large data sets, and multiple analysis. Identification of corresponding authors is critical for the transparency and clarity of any report. Without author names and affiliations, without identified offices in the MoH, the reader must ask, who is responsible to answer for this report? To whom must the public direct their questions and concerns about this report?
The WHO has simultaneously broadcasted and vanished from this report.
The described methods of this report are not without fatal shortcomings. First and foremost, an epidemiologic study must clearly show that individuals who were selected for the study accurately represent the population of interest. To that end, methods must offer clear and justifiable criteria for the inclusion of individuals in the study, and for their exclusion from the study.
The methodology section of this report simply declares that the selection criteria were “determined by the Ministry of Health”. The critical questions of “on what basis” and “why” remain unanswered. Selection criteria have major and critical influences on an epidemiological investigation and are universally expected to be fully discussed, even in short reports.
We cannot tell whether selection bias, a common problem in epidemiological studies, has occurred here. If it has, then the study is fully discredited. Based on information available in this report, we cannot rule out selection bias issues. The undisclosed criteria for recruitment of participants appears to have “included areas that had and had not been exposed to bombardment or heavy fighting.”
The maps and tables in the report do not indicate which areas were exposed to bombardment or heavy fighting and which areas were not. Another fatal shortcoming of the report is that the exposed and unexposed populations remain unidentified throughout. How is a comparison between two population’s rates of “spontaneous abortions”, “stillbirths”, or “congenital birth defects” possible if their exposure status is never described?
Using unscientific methods
The “Findings” section summarily reports a few basics of the study. These include the number of visited households and of completed questionnaires. Shockingly, the findings say that “around a quarter (27.5 percent) of the congenital birth defects were diagnosed by parents”. One may wonder, how reliable is a parent’s diagnosis of a newborn’s defect? Many birth defects are only detectable by trained professionals and some remain undiagnosed until months after birth. Relying on untrained diagnosis of birth defects introduces yet another serious source of error into this study. Indeed, this study leaves much room for potential error in measurement.
But the data presented does not relate birth outcomes to exposure to bombardment or heavy fighting …. The data also disregards the exposure status of the population.
The section continues by stating that, on the whole, 32 percent of the collected data corroborated with medical records. In other words, almost 70 percent of data on birth defects was provided by people with no medical expertise in the detection of birth defects.
Wouldn’t a systematic examination of hospital records in different districts and governorates provide a more reliable measure of potential increases in birth defects in the Iraqi population? In fact, this simple method of examining hospital records for the occurrence of birth defects in Iraq has successfully been used in some available studies, indicating increases in both birth defects and cancers in selected Iraqi hospitals when comparing medical records from before and after the 2003 invasion.
The main question this report is trying to answer is: “Compared to districts that were not exposed to bombardment, has there been an increase in rates of birth defects in districts that were exposed to bombardment or heavy fighting?”
But the data presented does not relate birth outcomes to exposure to bombardment or heavy fighting. This failure to integrate exposure data to outcome data makes it impossible to detect any changes over time, in a particular district, as a result of high exposures. The data also disregards the exposure status of the population. How can a potential change in adverse reproductive outcomes be evaluated when the exposure status of the population remains unknown?
Full disclosure needed
A more informative way to present this data would have been by district, and by exposure, before and after 2003. In other words, tables could have been constructed for individual districts indicating which districts had high exposure and which ones had low exposure to bombardment or heavy fighting. This presentation would be more likely to show changes over time in individual districts, based on the exposure status of the population. Indeed, some may argue that in a study such as this, even “district” may be too large a unit of analysis to render real effects visible.
Based on numerous limitations and uncertainties, some of which are indicated in the report itself, the conclusions of this report are overstated. According to the data provided by its unknown authors, they cannot legitimately make any conclusion regarding rates of birth defects in the governorates of Iraq. This amounts to an immense failure of this report in accomplishing what it set out to do: to detect changes in adverse reproductive outcomes before and after exposure to bombardment or heavy fighting in Iraqi population.
This anonymous report seems to be pointing a finger of accountability at the largely anonymous attendees of two WHO meetings.
These consist solely of an “expert meeting” in June, and an “expert peer review” meeting held for two days in July. Out of these meetings, only six people are named: two from the US Centers for Disease Control and Prevention, three from the UK, and one from Norway.
There is no indication that any of these people authored the September report. There is also no concrete contribution to that report which can be attributed to any of these “experts”.
Hence we must insist on full disclosure of the “damning evidence that there has been a rise in birth defects and cancers” in Iraq.
Dr Mozhgan Savabieasfahani, a native of Iran, is an environmental toxicologist based in Michigan. She is the author of over two dozen peer reviewed articles and the book, Pollution and Reproductive Damage (DVM 2009).