In the past the primary aim of clinical studies on female patients with uncomplicated acute cystitis (AC) was the eradication of bacteriuria at the test-of-cure visit and the clinical outcome was used as confirmation. In such a classical study [8
] for inclusion a positive culture was defined as isolation of a uropathogen in quantities ≥105
colony-forming units (CFU)/mL urine with pyuria, defined as ≥10 leukocytes/mm3
, and bacteriologic response was assessed as eradication
CFU/mL of original uropathogen), persistence
CFU/mL of original uropathogen), superinfection
CFU/mL of a uropathogen other than the original pathogen at any time during active therapy), and new infection
CFU/mL of a uropathogen other than the original pathogen at any time after the end of therapy).
In contrast, Stamm et al. [9
] had already shown that the traditional diagnostic criterion, ≥105
CFU/mL of midstream urine, has a very high degree of diagnostic specificity (99%) but a very low level of sensitivity (51%), which means that only 51% of symptomatic women with lower urinary tract infections (UTI) could be identified, whose bladder urine—obtained by suprapubic aspiration or by catheter—contained coliforms. The authors found the best diagnostic criterion to be ≥102
CFU/mL (sensitivity, 95%; specificity, 85%) and suggested that clinicians and microbiologists should alter their approach to the diagnosis and treatment of women with acute symptomatic coliform infection of the lower urinary tract. In a more recent study, Hooton et al. [10
] confirmed that colony counts of E. coli
as low as even 10 to 102
CFU/mL in midstream urine were sensitive and specific for the presence of E. coli
in catheter urine in symptomatic women. Therefore, it is not surprising that in the study of Henry et al. [8
] of the 469 patients not valid for efficacy, 90% (421) were excluded because no causative organism was isolated in predefined quantity (i.e., ≥105
CFU/mL) before treatment. This number was about the same as the 422 patients evaluable for efficacy. This consideration shows that in the past only a highly selected group of patients with acute lower UTI were eligible for clinical studies, although most of the excluded patients may have had about the same symptomatology also caused by acute uncomplicated lower UTI.
Because the traditional diagnostic criterion of ≥105
CFU/mL has a very low sensitivity, guidelines by the Infectious Diseases Society of America (IDSA) supported by the U.S. Food and Drug Administration (FDA) recommended to include also patients with a bacteriuria of ≥103
CFU/mL of a uropathogen with only little loss of sensitivity (about 80%), but greater specificity (about 90%) as compared to the recommendations of Stamm et al. [9
], because routine microbiological techniques can more reliably identify 103
CFU/mL than 102
]. On the other hand, to measure eradication of bacteriuria then became more difficult, because these microbiological techniques have at least an error probability of a decimal power. Therefore, many studies still used the much higher, but easier to handle threshold of 105
CFU/mL as demonstrated above [8
For the clinical inclusion in such a traditional study [8
] each patient had to have ≥2 signs or symptoms suggestive of an acute uncomplicated UTI (i.e., dysuria, frequency, urgency, suprapubic pain) with an onset of symptoms within 72 h of enrollment. In the study mentioned, the urinary frequency was the most common symptom (97.6%), followed by urgency (95.0%), dysuria (89.6%), and suprapubic pain (89.6%). Most patients in this study reported that the intensity of their symptoms was mild to moderate, although urinary urgency was severe in 37.4% of patients.
Since in actual clinical practice, and also supported by recent guidelines [11
], culture and susceptibility testing are not often performed in young to middle-aged women with acute uncomplicated UTI, the accurate diagnosis made only by the patient’s symptoms has become more important. Therefore the ACSS was developed and validated as a self-reporting questionnaire for diagnosing AC in female patients by assessing typical and differential symptoms, quality of life, and additional health conditions, which may play an important role in such a clinical setting [1
]. As an optimal threshold to predict AC with 89.3% (95% CI; 81.0–93.7%) sensitivity and 92.5% (95% CI; 86.9–97.0%) specificity, respectively, a total score of 6 points in the domain of typical symptoms could be established in the reevaluated ACSS [3
Classically clinical outcome is evaluated in such a study [8
] on the signs and symptoms of UTI (see above) as Cure
(disappearance of or improvement in signs and symptoms of the infection such that additional antimicrobial therapy was not required) and Failure
(no apparent response to therapy, persistence of signs and symptoms of infection, reappearance of signs and symptoms at or before the test-of-cure visit, or use of additional antimicrobial therapy for the current infection). Since in the past clinical outcome was only supportive to the microbiological response, more exact clinical outcome measures were not necessary.
Since even therapeutic strategies of AC are today investigated in controlled randomized trials (RCTs) with only symptomatic versus antibiotic treatment, e.g., ibuprofen versus fosfomycin trometamol [14
], reliable measures for the clinical outcome will become critical. The primary aim of such a study is now the patient-reported outcome and not the microbiological response. At least in a pilot study, it has been shown that persistent, but asymptomatic bacteriuria after successful symptomatic therapy does not necessarily trigger early recurrence [15
]. In another study on young women with recurrent UTI it also could be established that treatment of asymptomatic bacteriuria between symptomatic episodes may even be harmful and trigger more frequent symptomatic recurrences and as expected is associated with a higher prevalence of antibiotic-resistant strains [16
]. Therefore patient-reported outcome instruments need to be developed to differentiate more carefully between Success/Cure
to measure benefit or risk in medical product clinical trials as suggested by the FDA [19
Defining Success/Cure as complete disappearance of all signs and symptoms caused by the infection would be ideal, but this cannot realistically be measured. Although the symptoms in the “Typical” domain are of course typical for AC, some patients will have similar symptoms at least of mild severity caused by other reasons. Although it can be assumed, that such kind of symptoms was already present before the onset of infection, reporting of subjective symptoms depends also on cultural behavior and present psychological conditions. Therefore, also in the past disappearance as well as improvement were used to define Success/Cure. The question remains how much improvement is necessary to define Success.
In the present study 48 female patients diagnosed with AC according to clinical and laboratory assessment and scoring by means of the ACSS (part A) about typical symptoms, differential symptoms, quality of life and additional conditions, were treated and the patient-reported outcome was evaluated with part B of the ACSS.
Considering the results in Table 6
the optimal evaluation would be to define Success
only in patients with a score of 4 or less in the “Typical” domain (mode Nr. 4) with no specific score >1, which should be confirmed at the same time with a score of 3 or less in the “Quality of Life” domain with no specific score >1 (mode Nr. 7). Such a procedure includes a complete disappearance or at least such an improvement of the typical symptoms, that the interference on the specific items in the two domains is only mild at the same time. If there is an obvious discrepancy between the two domains (typical symptoms and quality of life), patients need to be fully assessed having in mind that (i) these symptoms are in fact typical for acute cystitis, but may also be found at least in part in patients with other diseases; and (ii) quality of life can also be altered not only by the symptoms of acute cystitis but also by underlying conditions of the patient.
The differentiation between Success
made only with the scores of the “Main Symptoms” showed that the success rate at a given visit may be overestimated at least in a few patients. Therefore, we recommended to define Success
by a score of 4 or less in the “Typical” domain or in the “Five Typical Symptoms” domain (mode 3) with no item more than 1 as standard, which then could be also confirmed by the scores obtained in the “Quality of Life” domain as outlined in Table 6
. The reason why the same scoring threshold can be used for mode 3 and 4 is explained by the fact, that all patients finally rated as Success had a score of 0 in the subscale “visible blood”. This finding may also have clinical significance in that way that a patient at follow-up (e.g., test of cure) with total scores suggesting Success but with a score of ≥1 in the subscale “visible blood” should be investigated more thoroughly whether the hematuria might have been caused by other underlying urological or nephrological diseases than by hemorrhagic cystitis.
According to this recommended evaluation, 37 (77.1%) patients would be rated as Success and 11 (22.9%) patients as Non-success in the present study. At the first glance, such a Success rate seems to be lower than reported in most of the clinical trials, but one has to consider that these patients were not treated uniformly according to the most effective strategy and a fixed follow-up visit was not scheduled as part of the study following practice guidelines. Therefore, it can be assumed that many patients treated successfully did not return for a follow-up visit because the purpose of this study was to evaluate the practicability of the ACSS in everyday practice.
The study has also shown that simple overall summary questions like in the “Dynamics” domain and used in many past clinical studies, but never successfully validated, are not sufficient to differentiate between Success and Non-success. There was, however, a significant correlation with the score reduction in the specific items of the “Typical” and “Quality of Life” domains and the corresponding subscales. Several potential reasons for the lack of clear differentiation between the Success and Non-success can be discussed. In the present study, the follow-up visit was on average scheduled earlier than the “test-of-cure” visit is usually scheduled, so relapsed patients may not be accounted for and more patients with resolution of symptoms of AC are included in the sample. There were only 10 out of 48 patients with higher values of the scores obtained from the “Dynamics” domain (scores equal to 2 or 3, none for 4), thus, small numbers and an uneven split may have led to the lack of differentiation between Success and Non-Success. A further reason might be, that the questions in the “Dynamics” domain were not precise enough and need to be improved. To even increase the applicability of the ACSS slight modifications of the questions are suggested as follows:
Score 0: Now I feel back to normal (All symptoms are gone);
Score 1: Now I feel much better (Most of the symptoms are gone);
Score 2: Now I feel only somewhat better (Only some of the symptoms are gone);
Score 3: Now, there are barely any changes (I have still about the same symptoms);
Score 4: Now, I feel worse (My condition is worse)
Nevertheless, the present ACSS appeared to be a suitable instrument for patient-reported outcome measures because a clear and well-balanced differentiation between Success and Non-success can be performed. A further advantage of the ACSS would be to use it as patient’s diary to measure the time of the symptoms declining or how many patients reach a certain predefined goal. Such an instrument may be especially useful in controlled RCTs to assess not only differences in clinical efficacy at certain predefined visits but to get results almost every day to determine the effectiveness of different therapeutic strategies.