The American Academy of Pediatrics' journal, Pediatrics, recently published a review of adverse events reported in teens (ages 12-21 years) using insulin pumps (Adolescent Use of Insulin and Patient-Controlled Analgesia Pump Technology: A 10-Year Food and Drug Administration Retrospective Study of Adverse Events). The media picked it up, and announced Insulin pumps linked to injuries, deaths in teens.
But the publication simply isn't worth a pile of beans.
First, it's a summary of two entirely different subjects: insulin pumps, and a different device, patient-controlled analgesic (PCA) pumps. Sorry, but that's like reporting on apples and grapefruits, both of which can go into salads. Or auto accidents and gunshot wounds, both of which occur in teens. Insulin pumps and PCA pumps simply aren't the same device, the patients who using them aren't the same, and the number of reports isn't: 1594 for insulin pumps and 53 for PCA pumps. I'll ignore the PCA data for the rest of this discussion.
Second, there's no comparison to three critical groups: (a) adults and children using the same devices, and (b) teens with the same disease (insulin-requiring diabetes) who aren't using pumps, and (c) teens who do not have diabetes.
Adults and children using the same devices inevitably must have had reported events, but there's no mention here of what events occurred in adults or in children, nor how they compared to those events occurring in teens. I'd suspect that the events in adults were similar, and that events in children were different, but without presenting and discussing the comparison data, we have no idea. For instance, in 2005, 381 events (76.4%) were patient injuries and 109 (21.8%) were device malfunctions; we don't have any information here about how the under-12 and over-21 groups did in 2005: were there a higher or lower percentage of patient injuries or device malfunctions, or the same ratios? Indeed, we don't even know if the absolute number of reported events in children or adults was higher or lower than the 1594 in teenagers, nor how the number of events compared to estimated number of teenaged pump users (which the pump manufacturers should have been able to supply to the authors). And another problem: many of the reports that are submitted to the FDA lack data about age of the patient, and there's no idea if the reports lacking age information occurred in teens or adults or children. It would have been helpful if there had been a tabulation of total number of events, broken down by age: children, teens, adults, unknown.
The missing comparison with teens with diabetes who don't use pumps is a much bigger problem. It is discussed by Kim in her SharePost, Isn't Diabetes the Challenge -- not the Pump? She correctly points out that "the study did not have a control group of teens on multiple daily injections. I question if the pump is dangerous or if the challenge of diabetes is especially trying during the adolescent years." For instance, the always worrisome issue of suicidality in teenagers: "Two apparently committed suicide with pumps. How many diabetic teens committed suicide with injections?" And again, we don't have any clue as to whether the data would show approximately equal rates of suicidality in non-pumping diabetic teens, and whether that rate is higher or lower than the rate in non-diabetic teens!