For the drug information team at Wolters Kluwer, we encountered one such challenge with our drug interaction content related to serotonin syndrome/serotonin toxicity. Spurred on by the needs of our users, we devoted a year to enhancing the evidence quality and drug interaction monographs related to serotonin syndrome in our solutions.
Here’s how we did it.
Background: A medication-related condition
Serotonin syndrome or serotonin toxicity (SS/ST) is a potentially life-threatening condition that results from excessive serotonergic stimulation within the central nervous system. Typically, patients present with the following symptoms:
- Mental-status changes (e.g., confusion, delirium, coma)
- Autonomic hyperactivity (e.g., hyperthermia, diaphoresis, hypertension, tachycardia)
- Neuromuscular abnormalities (e.g., hyperreflexia, clonus, tremor)
However, not all of these findings are consistently present in all patients. Manifestations and symptoms of SS/ST can range from mild to life-threatening.
Diagnostic criteria to aid clinicians in identifying this condition are available, but due to the non-specific symptoms of the condition, the broad range of symptom severity, and the overlap with other known diseases (e.g., neuroleptic malignant syndrome), the diagnosis of SS/ST remains a challenge for clinicians.
What is important for clinicians to recognize, is that SS/ST is not an idiopathic drug reaction. Rather, it is a predictable consequence of excessive serotonergic stimulation most commonly caused by medications.
Many drugs possess an ability to enhance central serotonergic activity. This occurs via several known mechanisms including:
- Inhibition of serotonin reuptake
- Increased release of stored serotonin
- Decreased serotonin metabolism
- Direct serotonin receptor stimulation
Sometimes this is an intended mechanism of action of the drug, and other times it is an additional pharmacologic effect.
The concomitant use of two or more of these agents, or use of a single agent at high or supratherapeutic doses, will increase the risk for the development of SS/ST.
The management and treatment of SS/ST involves early and prompt recognition of the condition, discontinuation of all serotonergic medications, and the provision of supportive care. That care could include control of agitation, autonomic instability, and hyperthermia, along with possible administration of serotonin receptor antagonists, like cyproheptadine.
The problem as we saw it
Drug interaction content related to SS/ST was always present in Wolters Kluwer drug information resources, including the Lexicomp drug interactions tool and the Medi-Span Drug Therapy Monitoring System (DTMS) database and Interactions API.
We heard from users that they appreciated the depth and specificity of information that DTMS/Medi-Span provided about these interactions and the breadth of coverage provided in Lexicomp. However, we also heard concerns from our users about not consistently seeing the same information in both resources. We recognized that we could improve our content related to SS/ST by taking the strengths of each database and providing consistent interaction coverage across both DTMS/Medi-Span and Lexicomp.
In addition, the team noted that neither system employed an evidence-based system driven by primary literature to categorize drugs or drug classes based on their risk for contributing to the development of SS/ST. Previous categorizations were not risk-stratified. They were based on product label warnings or other secondary/tertiary literature sources.
We recognized that we needed an evidenced-based system to categorize drugs based on their risk for contributing to the development of SS/ST. However, because no other outside published or online resources that we could find had this type of information, we decided to develop our own evidenced-based approach.
Users routinely want to know what specific clinical action(s) they need to take to manage a potential drug interaction, and they expect that the ratings of the drug interaction monographs (i.e., severity, risk rating, documentation level) reflect the data and risks associated with a particular drug pair. When our systems grouped all serotonergic agents in the same way (no risk stratification), we were only able to provide one clinical management recommendation and one set of monograph ratings.
By completing this project, we were able to provide more consistent interaction outputs and improve our evidenced-based recommendations and warnings. Our clinical management recommendations for SS/ST and monograph ratings can be more specific and tailored to the serotonergic risk with a particular drug pair.
That’s what we have been able to give our users.
The development process
The development of SS/ST is a rare event and it is not typically captured or seen in clinical trials or even in retrospective database analyses. Therefore, we had to rely on individual case reports, case series, and even spontaneous reporting databases to identify SS/ST reports associated with each individual drug or drug class. For each identified case report, we aimed to categorize the strength of association between use of the drug or drug class and the development of SS/ST. Case reports were divided into 4 possible categories:
Highly Suggestive of SS/ST:
- The drug in question is given at usual doses
- Development of symptoms is temporally associated with administration of the drug
- Development of SS/ST occurred when the drug was combined with no more than 1 other serotonergic agent
Suggestive of SS/ST:
- The drug in question is given at usual doses
- Development of symptoms is temporally associated with the administration of the drug
- Development of SS/ST occurred when the drug was combined with 2 or more serotonergic agents
Possibly Suggestive of SS/ST:
- The drug in question is given at usual doses
- Development of symptoms is temporally associated with the administration of the drug
- The role of the drug in the development of SS/ST is uncertain or pharmacokinetic interactions or patient comorbidity are possible confounding/modifying factors or the diagnosis of SS/ST is highly questionable
Non-Contributory Reports:
- The drug is given at higher than usual doses/overdose
- Or the drug is one of three or more precipitating drugs (i.e., initiated/dose increased just prior to the development of SS/ST)
- Or it is not a precipitating drug and has been given along with 1 or more serotonergic agents
Once we identified and classified all of the SS/ST reports associated with each potential serotonergic drug or drug class, we were able to group drugs into high risk, moderate risk, and low/indeterminate risk serotonergic agents based on the following criteria:
Serotonergic Agents (High Risk)
Non-selective antidepressant MAOIs, SSRIs, and SNRIs, plus any drug with the following:
- At least 5 patient reports highly suggestive of SS/ST associated with the drug or class of drugs
OR
- At least 7 patient reports highly suggestive or suggestive of SS/ST associated with the drug or class of drugs.
AND
- Serotonergic risk is supported by the known pharmacology of the drug or drug class.
AND
- Lack of a significant body of contradictory data suggesting a lack of serotonergic risk (e.g., routine therapeutic coadministration).
Serotonergic Agents (Moderate Risk)
- Drugs that do not qualify for inclusion in the high-risk group but have a primary mechanism of action that directly increases serotonergic activity (i.e., decreased reuptake, decreased metabolism, or direct receptor stimulation)
OR
- At least 8 patient reports of highly suggestive, suggestive, or possibly suggestive of SS/ST associated with the drug or class of drugs.
AND
- Serotonergic risk is supported by the known pharmacology of the drug or drug class.
Serotonergic Agents (Low/Indeterminate Risk)
Drugs that do not qualify for inclusion in the high-risk or moderate-risk group but have:
- Product labeling that warns of the possibility of SS/ST with the use of the drug either alone or when combined with other serotonergic agents.
OR
- At least four case reports of highly suggestive, suggestive, or possibly suggestive of SS/ST associated with the drug or class of drugs.
It should be noted that case reports classified as non-contributory are not included in the grouping assessment as they do not provide a reliable association between drug use and the development of SS/ST.
It took about a year to fully implement this content in Lexicomp. We spent one to two months developing the criteria, four to five months collecting the data and evidence for each potential serotonergic drug, applying the criteria and placing them into appropriate risk groups, and then an additional three months outlining the needed interactions and writing the monographs.
SS/ST data availability
Content reflective of these new SS/ST interaction groupings became available to Lexicomp users at the end of October 2019. These changes are being implemented more gradually in the DTMS/Medi-Span database, with the initial changes released in October 2019 and final implementation anticipated to occur in the first quarter of 2020.
Once the project is fully implemented, the interaction content will be aligned between systems. However, interactions that typically require no clinical action (i.e., interactions between our moderate risk and low/indeterminate risk groups) will only be present in Lexicomp to help prevent an excess of alerts that are not clearly actionable from triggering in the Medi-Span system to help combat “alert fatigue.”
Reports of SS/ST continue to be published, and our understanding of risks associated with various drugs continues to evolve. We routinely search the literature to identify new reports of SS/ST associated with medication use, and we aim to update our groupings of serotonergic agents annually to ensure our classifications remain up to date and reflect the most current evidence.
Dr. Carrie Nemerovski is Senior Clinical Content Specialist - Drug Interactions, Clinical Effectiveness, at Wolters Kluwer, Health. She previously worked as a clinical faculty member and clinical pharmacist at Wayne State University and Henry Ford Hospital in Detroit, MI. Dr. Nemerovski earned her PharmD from the University of Michigan College of Pharmacy and completed a PGY1 and PGY2 residency in cardiology at the University of Michigan Hospitals and Health Centers. She maintains an academic relationship with U of M, serving as an advanced pharmacy practice experience preceptor for PharmD students.
