Some of the most common rhetoric used against psychiatrists is that the drugs that psychiatrists prescribe are somehow more dangerous than other drugs. There are numerous problems with the argument including the fact that psychiatrists don't really influence what medications are approved by regulators and the majority of the so-called psychotropic medications (up to 80%) are prescribed by primary care physicians. There are the associated arguments that they are overprescribed and ineffective and I will address those at another time. What is the evidence about dangerousness? I have previously commented on the issue of whether or not the medications used by our field cause a person to become homicidal. I will restrict this post just to the issue of medication complications and whether or not regulatory action needs to be taken against a medication.
The areas of pharmacovigilance and pharmacoepidemiology offer some insights into the area of drug dangerousness, but at this point there are few good studies available in public access formats. One study done in Wales showed that over a 5 year period there were about 100,000 incident reports related to medications or about 9.7% of all patient safety incidents. The incidents resulted in severe harm or death in 822 patients (0.9%) of the medication related incidents. The majority of reports were skewed toward reports from hospitals (75%) as opposed to primary care clinics. Looking only at the severe and fatal outcomes by drug class (Table 8) 2/13 drugs could be classified as medications used to treat mental illness. Benzodiazepines and antipsychotics were ranked 6th and 12th respectively. The top 5 drugs starting with number one were opioids, antibiotics, warfarin, heparin, and insulin. Any physician working on hospital safety committees is aware of the number of complications due to anticoagulation. To add further context the total population of Wales in 2011 was about 3 million people, but the total prescriptions per drug class or the critical denominator to determine any true complication rate is unknown.
Unfortunately in the US, we have no complete systems for pharmacovigilance. We have a long standing data base that has been around for decades that was used primarily to monitor physicians prescribing practices for pharmaceutical companies. Pharmaceutical representatives would detail physicians (introduce product information) and this company would sell the information about whether the detailing resulted in increased prescriptions of that product. Occasionally data from this large data base makes it out into the medical literature, but there is a serious question about how well marketing information works for pharmacovigilance. There is publicly available data from the Centers for Disease Control (CDC) indicating that there are about 50,000 deaths per year attributable to medications. The majority of these deaths are accidental and intentional overdoses and there is no granularity to look at common severe side effects like anaphylaxis. A third source of data is proprietary databases from health care companies, hospitals, and government agencies. Those sources often lead to questions about the generalizability of the conclusions from those studies.
A potentially useful regulatory measure is the number of medications that have been identified as problematic in post marketing surveillance and removed from the market for safety reasons. The best review I could find on that topic is reference 2. The paper looks at market withdrawals of new molecular entities (NMEs) approved by the FDA between the years 1980 and 2009. Of a total of 740 NMEs during that period, 118 (15.9%) were discontinued. Twenty six drugs out of 118 were withdrawn due to safety reasons or a total of 3.5% of the original approvals. Nervous system drugs represented a total 104/740 approved drugs and a total of 6.7% of the discontinuations as a percentage of the approvals. Safety withdrawals were a total of 3 drugs or 2.9% of the total approvals in this therapeutic class. The bottom line is that a total of 1 drug used for psychiatric indications out of 740 NMEs in the last 3 decades was a medication was withdrawn for safety reasons.
The authors go on to provide a high degree of granularity with a complete list of all NMEs that were withdrawn for safety reasons and they are listed in Table 3. The three nervous system drugs listed are nomifensine (an antidepressant), levomethadyl acetate ( a drug used to treat opioid dependence), and pergolide mesylate (a drug used to treat Parkinson's Disease and restless leg syndrome). The study apparently does not look at the issue of drugs where the manufacturer voluntarily discontinued sales. As an example, Bristol Myers Squibb discontinued sales of Serzone (nefazodone) in 2003 due to a low incidence of hepatotoxicity with serious outcomes like liver failure and the need for transplantation. The conclusion of this article is that the majority of of drug discontinuations are due to commercial reasons and not safety. They noted a trend for decreasing NMEs over time and an associated decrease in drug discontinuations.
Part of the problem with the perception of drug dangerousness, especially with medications used for psychiatric indications seems to be the idea that they should be devoid of side effects. That is certainly the ideal scenario, but that is not the approach taken by regulatory bodies like the FDA. Like any regulatory body that depends on politicians for funding there will always be a variety of political influences. In some cases the bureaucratic structure may be prioritized over scientific review. The paper by Qureshi, et al is a good example of a certain threshold of severe side effects that may lead to drug discontinuations for those reasons, but any inspection of current approved medications and their rare but serious side effects shows that there are plenty of concerns out there for commonly prescribed drugs in all classes. The regulatory concern is that many of these medications are useful for the people who really need them. When any medication is applied over a population of people, there is a likelihood of rare but very serious side effects. That is not a reason to call the drug dangerous, especially if there are people who benefit from taking it. There is also a likelihood of common side effects that are less dangerous but adversely impact quality of life. It is also easy to see the problem politically. In other words there is some kind of conspiracy driving the prescription of some medications as opposed to others.
The reality is that the patient has a decision to make and as I have pointed out before, I have really never encountered a person (including myself as a patient) who takes that decision lightly. There are additional interpersonal and psychological factors. My personal bias as a physician is that the primary goal of treatment is minimal to no side effects and that tolerating side effects is a decision made by the patient but informed by the physician. It always needs to be balanced against any therapeutic effect of the medication.
George Dawson, MD, DFAPA:
1: Cousins DH, Gerrett D, Warner B. A review of medication incidents reported to the National Reporting and Learning System in England and Wales over 6 years (2005-2010). Br J Clin Pharmacol. 2012 Oct;74(4):597-604. doi:10.1111/j.1365-2125.2011.04166.x. Review. PubMed PMID: 22188210; PubMed Central PMCID: PMC3477327.
2: Qureshi ZP, Seoane-Vazquez E, Rodriguez-Monguio R, Stevenson KB, Szeinbach SL. Market withdrawal of new molecular entities approved in the United States from 1980 to 2009. Pharmacoepidemiol Drug Saf. 2011 Jul;20(7):772-7. doi: 10.1002/pds.2155. Epub 2011 May 14. PubMed PMID: 21574210