Keywords diagnosis - questionnaire - signs and symptoms - premenstrual syndrome - premenstrual
dysphoric disorder
Palavras-Chave diagnóstico - questionário - sinais e sintomas - síndrome pré-menstrual - transtorno
disfórico pré-menstrual
Introduction
Premenstrual syndrome (PMS) is defined as a recurring pattern of symptoms that occur
during the premenstrual phase and decline soon after the start of menses.[1 ] It is characterized by physical, affective, and behavioral symptoms that significantly
impair the daily lives of women, including work and personal activities, during the
luteal phase and spontaneously resolve within a few days of the onset of menstruation.[2 ]
About 50–80% of women report that they experience at least some PMS symptoms during
the premenstrual phase, and the condition occurs in approximately 30–40% of reproductive
females that report symptoms of PMS, who end up requiring some type of treatment or
better attention.[1 ]
[2 ]
[3 ]
[4 ]
[5 ]
[6 ]
[7 ]
[8 ]
Women with severe symptoms, including at least one disabling affective symptom, to
the extent of causing marked functional impairment are classified as having premenstrual
dysphoric disorder (PMDD). It is estimated that PMDD affects from 3–8% of menstruating
women.[9 ]
The symptoms of PMS can be very similar to those of other diseases, such as depression
and anxiety crisis, among others; therefore, an accurate prospective evaluation of
the symptoms is required to make the diagnosis. The International Society for Premenstrual
Disorders (ISPMD) recommends that the diagnosis is only confirmed after reviewing
the data recorded for 2 consecutive menstrual cycles.[10 ]
The use of structured questionnaires is established, and several validated diagnostic
techniques are available. The most accepted and widely used system is the daily record
of severity problems (DRSP), a prospective, self-administered questionnaire. To use
the DRSP as a diagnostic tool for PMS, it is necessary that the patient fills it for
at least two consecutive menstrual cycles. However, this requirement limits this practical
applicability in the day-to-day care of patients with premenstrual symptoms.[11 ]
[12 ]
[13 ]
The premenstrual symptoms screening tool (PSST) is a retrospective questionnaire completed
during clinical consultation with the patient. It is a screening tool used to identify
women who suffer from severe PMS/PMDD.[14 ] It is less time-consuming and more practical than two cycles of prospective charting.
However, the retrospective assessment of symptoms has limited value and requires validation
against an established prospective technique, such as the DRSP.[11 ]
Based on these considerations, the aim of this study was to validate the PSST in relation
to the DRSP for the diagnosis of PMS and PMDD.
Methods
The selected women read and signed the informed consent, which was approved by the
Committee of Ethics in Research of the institution (number 15–0087).
This was a cross-sectional study with participants recruited by local media (TV, radio
or newspaper) announcements. The study was developed in a hospital in the state of
Rio Grande do Sul, Brazil, during the period from August of 2014 to December of 2015.
The participants were women between 20 and 45 years old with PMS complaints and cyclical
menses. The exclusion criteria were: menopause, use of any continuous hormonal contraception,
use of antidepressants or depression diagnosis.
It was determined that in order to get a Kappa coefficient of at least 0.7, with 80%
power, and a level of significance of 5%, the sample size should comprise at least
123 valid questionnaires. The women included were evaluated in terms of weight, height
and body mass index (BMI), and they all answered the primary care evaluation of mental
disorders (PRIME-MD) questionnaire (Module mood). In the cases of positive screening
for depression, the diagnostic and statistical manual of mental disorders (DSM-V)
questionnaire was applied to establish the diagnosis. The depressive women were excluded
from the study and referred for psychiatric care. After the PSST ([Supplemental Material 1 ]) was completed by the women, the DRSP ([Supplemental Material 2 ]) was handed out to be filled out daily for two consecutive menstrual cycles.[12 ]
[13 ]
[14 ]
[15 ]
[16 ]
[17 ]
[18 ]
Statistical analysis was performed using the software SPSS, version 18.0 (SPSS Inc.,
Chicago, IL, USA). Continuous variables were expressed as mean ± standard deviation
(SD) and categorical variables were expressed as absolute (n ) and relative (n %) frequencies. The agreement between the two questionnaires was evaluated by calculating
the Kappa (k) and the prevalence-adjusted, bias-adjusted kappa (PABAK) values. Kappa
is an agreement measure beyond that agreement expected to occur by chance. The Kappa
expected, according to Altman,[19 ] was 0.8— with a range of 0.643–0.893, with a 95% confidence interval (95%CI). Additionally,
an interval-by-interval basis was calculated by the PABAK.[20 ] The PABAK is a simple and flexible index, representing a method for calculating
inter-rater reliability between two raters (for example, DRSP and PSST) using an ordinal
rating scale of categories (such as, Normal, PMS and PMDD), suitable for nominal scales
only. The PABAK value was calculated using the computer program for Epidemiologists
WinPEPI, version 11.65 (for 2 × 2 categories tables), or the PABAK-OS calculator,
available at singlecaseresearch.org (for table formats other than 2 × 2 categories).
For all analyses, significance was set at 5%.
Results
Overall, 377 women signed up to participate in the study; however, only 282 (74%)
of them were considered eligible due to category recruitment and thus, these women
answered the PSST. The baseline characteristics of these women are shown in [Table 1 ].
Table 1
Sociodemographic characteristics of the women who completed the PSST and filled out
the DRSP for 2 consecutive menstrual cycles
Variables
(n = 127)
Age (years) - Mean ± SD [Min-Max]
33.7 ± 6.6 [20.0–45.0]
Skin color - n (n %)
White
112 (88.2)
Not white
15 (11.8)
Educational level - n (n %)
Elementary
3 (2.4)
High school
37 (29.1)
Academic
85 (66.9)
Using non-continuous regimens of combined hormonal contraceptive - n (n %)
Yes
72 (56.7)
No
55 (43.3)
Smoker - n (n %)
Yes
4 (3.1)
No
123 (96.9)
BMI - n (n %)
Normal
59 (46.5)
Overweight
44 (34.6)
Obese
24 (18.9)
Abbreviation: BMI, body mass index; SD, standard deviation.
Data expressed as mean ± standard deviation (SD) or absolute and relative frequencies
[n (%)].
Of the eligible women, 135 (55%) did not fully complete all the two cycles of the
daily questionnaire (DRSP). Only 127 (45%) women completed the study— including the
2 cycles of the DRSP. These women were considered to have fully participated in this
study and represent the full analysis set ([Fig. 1 ]).
Fig. 1 The flowchart shows the selection process of the participants.
The Portuguese version of the DRSP was applied, and an initial exploration for the
discriminant validity of this questionnaire was performed, proving the efficiency
of its use for the diagnosis of PMS/PMDD. Confirming the diagnosis of PMS, the [Supplemental Material 3 ] presents the medians of the DRSP scores for individual questions considering the
luteal and follicular phases. The mean follicular phase scores were ≤ 30% of the luteal
phase scores for at least 1 out of 3 different symptoms. Impressively, all of the
DRSP items were significantly higher in the luteal phase when compared with the follicular
phase, proving the sensitivity of this questionnaire for this purpose. These results
support the validation of the DRSP instrument in Portuguese for the Brazilian population.
[Fig. 2 ] shows that the diagnosis of PMS was higher with the DRSP, and the PMDD diagnosis
was higher using the PSST. The percentage of PMS diagnosed by the DRSP was 74.8%,
and 41.7% by the PSST; the percentage of PMDD by the DRSP was 3.9%, and 34.6% by the
PSST. The number of patients that were considered “normal” was similar in both questionnaires.
Fig. 2 Comparative prevalence (percentage of total sample) of PMS/PMDD diagnosis between
the PSST and the DRSP. Both instruments presented similar percentages for values falling
within normal range. The PMS percentage was 74.8 by DRSP and 41.7 by PSST. The PMDD
percentage was 3.9 by DRSP and 34.6 by PSST. Abbreviations: DRSP, daily record of
severity problems; PMDD, premenstrual dysphoric disorder; PMS, premenstrual syndrome;
PSST, premenstrual symptoms screening tool.
When we compare the diagnosis in both questionnaires, 70% of the women considered
“normal” by PSST had a PMS diagnosis after the DRSP and 83% of the women with PMS
diagnosis by PSST had the PMS diagnosis confirmed by the DRSP. Otherwise, only 6.8%
of the PMDD diagnosis by the PSST had a PMDD diagnosis by the DRSP, and 68.2% of the
PMDD diagnosis by the PSST had a PMS diagnosis by the DRSP ([Fig. 3 ]).
Fig. 3 Assessment of the PSST results according to DRSP diagnosis. Normal percentage by
PSST (70%, light gray column in normal) would be classified as PMS by DRSP diagnosis.
Premenstrual syndrome percentage by PSST (83%, light gray column in PMS) with confirmed
PMS diagnosis by DRSP. Premenstrual dysphoric disorder percentage by PSST (6.8%, darker
gray column in PMDD) with confirmed PMDD diagnosis by DRSP. Premenstrual dysphoric
disorder percentage by PSST (68.2%, light gray column in PMDD) would be classified
as PMS by DRSP diagnosis. Abbreviations: DRSP, daily record of severity problems;
PMDD, premenstrual dysphoric disorder; PMS, premenstrual syndrome; PSST, premenstrual
symptoms screening tool.
When considering all the frequencies in the diagnosis categories (Normal, PMS and
PMDD) between the DRSP and PSST, an agreement between both instruments could not be
observed, as illustrated by the value of the Kappa coefficient (k = 0.104, 95%CI [0.001–0.207],
p = 0.039). Even when considering the adjusted prevalence through the PABAK coefficient,
its value (0.161, 95%CI [0.079–0.243] for all diagnosis categories (Normal, PMS and
PMDD) is still poorly agreed on between the DRSP and PSST [data not shown ]. Additionally, when considering both coefficients comparing the normal and PMS/PMDD
categories, the PABAK coefficient value (0.39) showed a moderate consistency between
the DRSP and the PSST, although it still fits into a low intensity, according to Altman
(1990).[19 ] The results of “normal versus PMS/PMDD diagnosis” and “PMS versus PMDD diagnosis” between the PSST and DRSP are displayed in [Table 2 ].
Table 2
Evaluation of PMS/PMDD diagnosis using the PSST
Variables
Normal versus PMS/PMDD diagnosis
PMS versus PMDD diagnosis
Sensitivity [95%CI]
79.00% [70.02%–85.83%]
60.00% [23.07%–88.24%]
Specificity [95%CI]
33.33% [18.64%–52.18%]
59.46% [48.08%–69.91%]
Positive Predictive value [95%CI]
81.40% [76.68%–85.42%]
9.1% [4.34%–18.06%]
Negative Predictive value [95%CI]
30.00% [18.32%–45.02%]
95.7% [88.46%–98.44%]
Kappa [95%CI]
0.119 [−0.068–0.305]
0.054 [−0.076–0.184]
p
0.181
0.393
PABAK
0.39
0.19
Abbreviations: CI, confidence interval; PMDD, premenstrual dysphoric disorder; PMS,
premenstrual syndrome.
Discussion
Experts diverge about the most suitable PMS diagnostic tool. Many group members favored
the DRSP; however, more simple methods are desirable for clinical use and for screening
patients for research studies.[11 ]
The challenge for the diagnosis and classification of PMS is to distinguish women
who need treatment from those without clinical relevance.[21 ] The PSST is a fast and effective screening tool and an important starting point
for further assessment.[14 ]
When comparing the data of the PSST with the results of the DRSP, our study shows
that the PSST overestimated the diagnosis of PMDD and minimized the diagnosis of PMS.
This may be explained by the fact that the PSST data was not collected over time.[14 ] However, it is a good screening tool for PMS with a high sensitivity (79%).
This study points to an important conclusion: we had many losses; 135 women abandoned
the recordkeeping, citing the amount of time taken to fulfill the record as a reason
for the abandonment. The problem is that filling out the DRSP is a time-consuming
process, which is frequently not followed through. An epidemiological study showed
that 30% of women refused to participate in the study because they did not want to
fill out a daily record. The recordkeeping in question may even cause resistance to
future treatment offerings. Therefore, it is important to develop economic diagnostic
tools for clinical practice.[21 ]
In a study with 1,477 women, only 56% completed the DRSP. This study included Brazilian
women and showed an elevated incidence of moderate to severe PMS/PMDD: 47.7%.[9 ] This demonstrates how difficult it is for women to keep a daily record as part of
a research. In the clinical practice, the use of PMS/PMDD diaries is even less frequent.
This study has some limitations. Although 282 women completed the PSST, only 127 women
returned with 2 cycles of DRSP filled out. This highlights the difficulty in the application
of the DRSP in clinical practice and research, suggesting that the PSST should be
used as the initial screening tool.
Conclusion
Considering that PMS/PMDD have a well-validated diagnostic tool, the DRSP, the PSST
should be considered as a screening tool. The PSST underreported the PMS diagnosis
and over-estimated the PMDD diagnosis. We recommend that positive cases of PMS/PMDD
diagnosed by PSST are further evaluated by using the DRSP.
