Writing the side-effects section of a consent form
The side-effects section of a consent form is selected, ordered and worded so that a lay reader can find in it what matters to their own safety.
Selecting the side effects
Drugs can be associated with a wide variety of side effects; one study found an average of seventy on prescription drug labels. As a result, patients may experience information overload and may overlook information that is important to their safety. The side-effects section of a consent form must be carefully balanced to be as understandable as possible to lay readers while being sufficiently comprehensive.
Writing and editing the side-effect language of a consent form is a collaboration among principal investigators, scientific editors, study sponsors, ethics committee members and members of the lay community. At MD Anderson Cancer Center, an ad-hoc committee of the ethics committee, made up of doctors, research staff and community members, verifies the accuracy of every lay translation of a side effect, with a focus on what the patient may experience. The same committee decides whether each side effect counts as serious under the definition of the US Food and Drug Administration (FDA): fatal, life-threatening, causing hospitalisation, disability, permanent damage or birth defects, or requiring intervention to prevent permanent impairment.
A side effect the committee classes as nonserious that occurs rarely, less than 3% of the time, can be left out of the drug listing. Many drugs have numerous side effects that appear in post-marketing studies but only rarely, so leaving out the rare, nonserious ones shortens the form substantially and emphasises the more critical information. The length of the form is one of the primary obstacles to a participant retaining what it says, and the shorter the form, the better participants retain it.
Combining similar side effects
Similar side effects can be combined. Where a drug listing carries sloughing, desquamation, skin exfoliation and local exfoliation, the four become skin peeling. Duplicate terms that are too similar are removed: shortness of breath does not need to appear alongside difficulty breathing, nor drowsiness alongside grogginess. Removing duplicates makes the form shorter, so that it can hold the reader's attention, which leads to better comprehension and retention. But it might be misleading to list bleeding gums without bleeding in the brain, and combining the two into a single entry for bleeding would take away from the information that potential participants should be given.
Lay wording of each entry
Some side effects, such as loss of appetite and difficulty sleeping, are simple to convey, and others are obscure. For an obscure side effect, it is important to include its possible signs and symptoms, so that the patient can better identify an adverse reaction when it happens. Reversible posterior leukoencephalopathy syndrome becomes brain damage that may be reversible, with possible headache, confusion, seizures or vision loss, and a low sodium level becomes a low blood level of sodium, with possible headache, confusion, seizures or coma.
A generic description of what the patient may experience is used instead of technical jargon for the actual problem. Palpitation, premature ventricular beats, prolonged QT interval and AST wave changes have precise meanings to experts, but for lay readers they are all written as irregular heartbeat. In the same way, amnesia becomes memory loss, pyrexia becomes fever, Stevens-Johnson syndrome becomes very severe blistering skin disease with ulcers of the skin and digestive system, and increased partial thromboplastin time becomes increased risk of bleeding.
Ordering by likelihood and body system
For a drug the FDA has approved, MD Anderson's scientific editors compile every side effect it is known to cause from the centre's formulary. Where the likelihood is known, they sort the side effects into three bands: likely, in more than 20% of patients; common, in 3% to 20%; and rare but serious, in fewer than 3%. This highlights the most common side effects by moving them to the top. Within each band the side effects are grouped by body system, such as cardiac, dermatologic and respiratory, so that editors, investigators, research staff and, above all, patients can find a particular one quickly. Most are listed as concise bullets in a three-column chart, which holds down the page count, and certain side effects that need more description and do not fit easily as bullets, such as myelosuppression, are written as paragraphs outside the chart.
For a drug the FDA has not approved, the editors take the possible side effects from the investigator's brochure and the sponsor's model consent form. For many such drugs it is too early to know how often each side effect occurs, and banding them by likelihood would be misleading because of the low statistical power. All known side effects are then listed under a heading stating that their frequency is unknown.
The standing preamble
At MD Anderson, the side-effects section of almost every clinical consent document opens with the same standard paragraph, which lets the bulleted side effects beneath it be listed without repetitive descriptors such as serious and fatal. The paragraph says that the participant is at risk for side effects while on the study and that side effects vary from person to person. It says that the more common side effects and the rare but serious ones are listed, that the reader should discuss them with the study doctor, and that the reader may want to ask about uncommon side effects seen in small numbers of patients but not listed. It ends by saying that many side effects go away shortly after treatment stops, but that some may be serious, long-lasting or permanent and may even result in hospitalisation or death.