3.1. Initial design
-
The literature review identified five randomized trials comparing antibiotic versus placebo in uUTI [5, 29–32]. These five trials were all included in a meta-analysis published in 2009 [6].
We selected four randomized trials comparing the use of an antibiotic to an non-steroidal anti-inflammatory drug (NSAID) in uUTI [18] [19] [33] [34] and four studies describing the natural history of uUTI simple [10, 35–37].
We selected four examples of available PtDAs. Three PtDAs concerned anti-infectious treatment decisions in outpatient settings [38–40]. A fourth PtDA was blank and could be used for any shared decision-making situation [41].
Symptom duration
Untreated uUTI healed spontaneously in 50–70% of cases. Symptoms could last up to several weeks [36]. Mild to severe symptoms improved after 4.94 days in women not taking antibiotics [37]. Symptoms in women not taking an antibiotic lasted 50–60% longer than in women treated with an antibiotic to which the bacterium was susceptible [37]. Clinical resolution of symptoms was more likely in patients treated with antibiotics, with an odds ratio of 4.67 [2.34–9.35] [6].
After three days, the proportion of complete resolution of symptoms varied across studies, from 37% in patients treated with nitrofurantoin versus 20% in patients treated with placebo [5], 44% in women treated with fosfomycin versus 24% in women treated with ibuprofen [18], and 80% in women treated with norfloxacin versus 54% in women treated with diclofenac [34].
Risk of complications
There was no significant difference associated with the risk of pyelonephritis when compared between patients taking an antibiotic and those taking a placebo (OR 0.33; CI [0.04–2.70]). The incidence of pyelonephritis was between 0 and 2.6% [6]. There were no reported cases of sepsis. The French guidelines describe the risk of pyelonephritis as very low [9].
Three of the four trials comparing antibiotic use with an NSAID in uUTI found more pyelonephritis in women taking an NSAID compared to women taking an antibiotic [18] [19] [34].
Adverse reactions
The occurrence of adverse events was significantly higher in antibiotic-treated patients compared to placebo-treated patients [6].
In the case of pivmecillinam, 5–8% of adverse events were reported [32]. In a multinational trial conducted in primary care and hospital settings, patients taking single dose fosfomycin had 6% adverse events versus 8% of patients taking nitrofurantoin. The most common adverse events were gastrointestinal (nausea, vomiting, diarrhea, abdominal pain), asthenia, headache, dizziness, and vaginal discharge [5] [42]. These studies did not report any serious allergic reactions related to antibiotics.
Recurrence
The data did not allow for a meta-analysis on the occurrence of clinical recurrence [6]. In the study comparing nitrofurantoin to placebo the clinical recurrence rate at two weeks was between 17.6–20% [5]. In the study comparing pivmecillinam to placebo the recurrence rate at one month was 12–13% [32]. We did not find study comparing the incidence of recurrence over longer time periods.
Resistance
The emergence of resistance in the randomized studies varied from 0–45.5% of women taking an antibiotic versus 0–20% of women taking a placebo, with no significant difference [6]. In a Swedish study, taking an antibiotic for a uUTI in primary care was associated with a higher rate of bacterial resistance [13].
Alternative treatments
Patients taking herbal medicine do not have a different symptom course than those taking a placebo [43–44]. There is no evidence of cranberry (Vaccinium macrocarpon) or hydration as treatment for cystitis simple [45–46].
3.1.2. Focus groups
Participants spoke of their personal or reported experiences with cystitis and its impact on their social and sexual lives: "It restricts social life, because you always have to be near a toilette (laughter)" (P3.2), "You don't dare to have sex anymore" (P4.4).
This experience touched on intimacy and could be perceived as taboo: "It's a feeling of guilt, actually. Well, in a way it is, because we feel that our intimacy as women is being attacked" (P3.4).
They feared that cystitis could be complicated by renal, gynaecological, or fertility problems: "If there's blood in my urine it means that the kidneys must be affected" (P3.2), "It's going to make an infection maybe a bit generalized in that area, maybe causing problems to have children..." (P1.2).
Their knowledge about the risk factors and treatment of cystitis was part of a lay knowledge shared amongst women: "I told my mother about it and she said: don't worry, drink lots of water, it will pass, this antibiotic works well; because she often had it" (P2.2).
Some participants described a feeling of infantilization and guilt during the consultation with the physician: "The doctor or the ones I saw, made me feel like it was my fault because I didn't wash (myself) well. Afterwards, we are told once we hold it back! We don't have to be told every time" (P3.2).
They wanted a personalized exchange, where they could express their experiences: "What is important with cystitis, I learned from the doctor who took the time to explain it to me. [...] We are not in a normal state when we are sick. So he really needs to listen to us" (P2.5).
The participants wished to clearly define cystitis and its risk factors, with a vocabulary accessible to all, without medical jargon: "And in rather simple terms, so that everyone can understand it... Not in doctor's language" (P1.3). They suggested a pictorial presentation: "The more graphic, the more people are affected" (P1.3), accompanied by the doctor: "The diagram is nice, but if the doctor doesn't explain it to you, [...] she won't understand anything" (P3.3).
Their expectations of treatment could be the rapid relief of symptoms, or the prevention of recurrence in the longer term: "Isn't there something more effective and long-lasting, [...] rather than just immediately stopping the pain?" (P1.2).
Some patients have expressed an interest in being involved in the decision related to the antibiotic: "Do you have something to offer me that is not antibiotics? I have time now, I can stay at home, if it's really not going well we'll switch to antibiotics, but why don't we test something else? Maybe there should be a second option" (P4.4).
The action of the antibiotic was seen as magical, but could lead to side effects and resistance: "This antibiotic was really a miracle" (P6.1), "Every antibiotic [...] that we swallow, we know that there are side-effects" (P1.1).
The participants considered alternative treatments, described as natural, such as cranberry or hydration: "Having the choice between a chemical molecule and something a bit more natural, something less harsh, I'll take what is less harsh" (P2.3).
3.2. Draft version of prototype
The draft version of the prototype (Fig. 2) included the following elements.
The title explicitly described the decision whether to take an antibiotic or not. The elements of the PtDA were then arranged according to these two choices.
The common symptoms and etiology of uUTI were briefly described and illustrated with a diagram of the bladder.
The treatment options that were presented included antibiotic treatment, hydration, and cranberry. The practical modality of a single-dose antibiotic therapy was specified.
The average duration of symptoms, the risks of recurrence, resistance and complications were detailed according to whether the antibiotic was taken or not.
Colored pictograms numerically represented the evolution of symptoms after three days according to the choice of treatment and the incidence of adverse effects.
The patient values to be explained were physical discomfort, the impact on their daily life (professional, social, sexual), their general opinion on antibiotics and their adverse effects. A free space allowed for the collection of additional value.
Deliberation was facilitated by sliders polarized according to the two options, for each value expressed. A final slider helped in the decision-making process.
The chosen format of the PtDA was a double-sided A4 sheet of paper. It was intended to be used during discussion with the physician during the consultation and not for use by the patient alone.