Comparison of Two Methods for Assessing Erectile Function Before Radical Prostatectomy

Abstract

Patient-reported outcome instruments for erectile function often ask respondents about their experience over the previous 4 wk. This is problematic for baseline assessment of patients with prostate cancer (PC) before treatment, as the previous 4 wk would probably have involved procedures such as biopsy and considerable anxiety related to their diagnosis. At San Raffaele Hospital, the International Index of Erectile Function (IIEF-6) was used to ask new PC patients about function in both the previous 4 wk and 6 mo. We compared responses to these two timeframes. IIEF-6 scores were lower for the 4-wk period (median 24 vs 26; p < 0.0001) predominately because approximately one in six of patients with good function in the 6-mo time frame had very poor function in the 4 wk before completing the questionnaire (adequate erectile function 60% and 51%; absolute difference 9%, 95% confidence interval 8–10%). Results were further confirmed using a comparison group of 5395 patients with PC newly diagnosed at Memorial Sloan Kettering Cancer Center who had similar function in the previous 6 mo. Erectile function evaluation for men presenting with PC should involve asking about typical function over a 6-mo period rather than focusing on the previous 4 wk.

Patient summary:

Questionnaires to assess erectile function often ask men about function in the previous 4 wk. We found that this underestimates function in new prostate cancer patients and that such men should be asked about typical function over a 6-mo period.

Many patient-reported outcome instruments for erectile function—including the International Index of Erectile Function (IIEF) and the Expanded Prostate Cancer Index Composite (EPIC)—ask about a man’s experience in the previous 4 wk. This is problematic for baseline assessment of prostate cancer patients before treatment. The 4 wk before such a baseline assessment may have included prostate biopsy, an uncomfortable procedure that probably interferes with a man’s interest in sex [1,2], and news of a cancer diagnosis, which causes anxiety and thereby influences sexual function [3]. Thus, assessment of typical function before radical prostatectomy may not be reliable if it asks about the previous 4 wk. At San Raffaele Hospital, when the IIEF-6 is used to evaluate baseline erectile function for new prostate cancer patients, we alter the language. The day before radical prostatectomy, patients complete two questionnaires for assessment of typical erectile function: the standard questionnaire, which evaluates erectile function in the previous 4 wk, and a second questionnaire that asks the same questions about the previous 6 mo.

Here we compare these two approaches for evaluating erectile function. For each patient, we compared the IIEF-6 scores between the 4-wk and 6-mo timeframes. Among San Raffaele patients, 352 men had a missing IIEF-6 score for the previous 6 mo and were excluded from the analyses. We performed confirmatory analyses using a large cohort of newly diagnosed prostate cancer patients evaluated at Memorial Sloan Kettering Cancer Center (MSKCC) who were asked about typical function in the previous 6 mo. For these patients, the standard language in the IIEF-6 about function in the previous 4 wk is substituted by: “Being diagnosed with cancer can be very stressful, and this sort of stress can often cause other health problems, such as a lack of interest in sex. When you fill in the questionnaire, your doctor will want to know how you have been doing in general over the past six months.” Patients are given the questionnaire as part of their preoperative care, usually within 30 d before radical prostatectomy.

Among 2369 Milan patients, the IIEF-6 score was lower for the 4-wk than the 6-mo timeframe before surgery (median 24 vs 26; absolute difference 2, 95% confidence interval [CI] 2–3; p < 0.0001). Figure 1 is a scatterplot of scores for the two timeframes. It is clear that the results are driven by a group of patients with good erectile function over the 6-mo period but very low function in the 4 wk before surgery. The proportion of patients who had a preoperative IIEF-6 score ≥24—one cutoff that is widely used for adequate erectile function [4]—was 60% for the 6-mo and 51% for the 4-wk timeframe (absolute difference 9%; 95% CI 8–10%). More specifically, 254 patients (11%) who had erectile function in the 6-mo timeframe met the criterion for erectile dysfunction for the 4-wk timeframe, with a median decrease in score of 11 (interquartile range 4–21).

Fig. 1 –

Preoperative International Index of Erectile Function (IIEF) scores for 2369 patients treated at San Raffaele Hospital for the 4-wk and 6-mo timeframes targeted by the questionnaire. Points are jittered for illustrative purposes. The black 45° line acts as a reference. The dashed lines indicate one widely used cutoff for adequate erectile function.

In the confirmatory analysis, preoperative IIEF-6 score for the 6-mo timeframe was not different between the MSKCC and San Raffaele patients (Table 1), with a difference of −0.43 points (95% CI −0.91 to 0.05; p = 0.08) after adjustment for age and comorbidities. There are three possible explanations for our results: (1) a true decrease in function over time; (2) overestimation of previous erectile function, a form of optimistic recall bias; and (3) high anxiety and medical procedures have a negative effect on erectile function shortly before surgery. We believe that the data strongly support the third hypothesis. This is because the findings are driven by a subgroup of men who reported good function over the previous 6 mo but little or no function in the previous 4 wk. A true decrease in organic function seems unlikely, given that the natural history of erectile function loss with aging is well understood and does not involve sudden loss of function over a short timeframe. It also seems implausible that a man reporting an IIEF-6 score of, for example, 3 in the previous 4 wk would report near-perfect erectile function in the previous 6 mo if this was not in fact the case. Therefore, although we cannot definitively identify causal factors, the most reasonable explanation for our findings seems a change in erectile function shortly before surgery. Our study is not devoid of limitations. Data on comorbidities, number of prior biopsies, changes in the use of medications or other medical aids—all potentially affecting potency—were not available for the analyses. However, since our aim was not to examine causality (ie, determinants of lower erectile function in the previous 4 wk), such factors do not represent confounders and thus they probably had little, if any, influence on our results. It is possible that the reliability of recall might be time-dependent; that is, a patient might be more prone to remember a problem with his erectile function in the previous 4 wk than in the previous 6 mo. However, there is independent evidence that patients are able to recall erectile function accurately [5]. We note also that the Sexual Health Inventory for Men [6], an instrument that is widely used in clinical practice, uses a 6-mo timeframe. This gives us reason to believe that a 6-mo recall period is considered reasonable by urologists and that the wording of other questionnaires, such as the IIEF and EPIC, can and should be changed to 6 mo when assessing baseline erectile function for men newly diagnosed with prostate cancer.

Table 1 –

Baseline characteristics for 7764 prostate cancer patients undergoing radical prostatectomy at two high-volume institutions from 2008 to 2018

	San Raffaele	MSKCC	p value a
Patients, N (%)	2369 (31)	5395 (69)
Median age, yr (IQR)	64 (58–69)	62 (57–67)	<0.0001
Smoking habit, n (%) [N = 6849]
Never	1145 (49)	2405 (53)	0.002
Current	291 (12)	499 (11)
Former	908 (39)	1601 (36)
Hypertension, n (%)	1025 (43)	2591 (48)	0.0001
Diabetes, n (%)	163 (6.9)	596 (11)	<0.0001
Hypercholesterolemia, n (%)	450 (19)	2744 (51)	<0.0001
Median preoperative IIEF-6 score (IQR)
For previous 6 mo	26 (17–29)	26 (16–30)	<0.0001
For previous 4 wk	24 (9–29)	NA
pIIEF-6 score ≥24, n (%)
For previous 6 mo	1410 (60)	3157 (59)	0.4
For previous 4 wk	1197 (51)	NA
Patients with pIIEF-6 score of ≥24 for previous 6 mo and <24 for previous 4 wk, n (%)	254 (11) b
Median pIIEF-6 score for previous 6 mo (IQR)	27 (25–28)
Median pIIEF-6 score for previous 4 wk (IQR)	16 (5–22)
Median difference in pIIEF-6 score (IQR)	11 (4–21)

pIIEF = preoperative International Index of Erectile Function; IQR = interquartile range; MSKCC = Memorial Sloan Kettering Cancer Center; NA = not applicable.

^aDifferences between groups were assessed using the Wilcoxon rank-sum and χ² tests for continuous and categorical variables, respectively.

^bCharacteristics for the 41 patients who had an IIEF-6 score of <24 for the previous 6 mo and ≥24 for the previous 4 wk are presented in Supplementary Table 1.

In conclusion, the wording for the timeframe in questions affects the reliability of erectile function evaluation for men with newly diagnosed prostate cancer. Use of a 4-wk timeframe lowers apparent baseline function for some men, probably because of anxiety and sequelae of procedures such as biopsy. This underestimation is concerning and might result in wrong counselling and treatment decisions (ie, indication for nerve-sparing surgery). Patient-reported outcome instruments should use a 6-mo timeframe for assessing erectile function, and by the same token, other aspects of sexuality such as libido and bother for this group of patients.

Evaluation of erectile function on presentation with prostate cancer is more accurate if patients are asked about function over the previous 6 mo rather than the previous 4 wk. Patient-reported outcome instruments should be modified appropriately for this population.