The European Journal of Contraception and Reproductive Health Care, 2014; Early Online: 1–9

A new method to detect significant

basal body temperature changes
during a woman’s menstrual cycle
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14

Günter Freundl , Petra Frank-Herrmann , Simon Brown† and Leonard Blackwell ‡

Natural Fertility Section, Universities of Heidelberg and Düsseldorf, Germany, †Deviot Institute, Deviot, and School
of Human Life Sciences, University of Tasmania, Tasmania, Australia, and ‡Institute of Fundamental Sciences, Massey
University, Palmerston North, New Zealand

............................................................................................................................................................................................................

ABSTRACT Objective To compare the results of a computer programme based on the Trigg’s tracking
system (TTS) identification of the basal body temperature (BBT) shift day from daily records
of BBT values (TTS transition day), with the BBT shift day identified from the same records
using the Sensiplan® symptothermal method of natural family planning.
Methods A computer programme was written to display the daily BBT readings for 364
menstrual cycles from 51 women aged 24 to 35 years, obtained from the German Natural
For personal use only.

Family Planning (NFP) database. The TTS transition day so identified from each record was
then compared with the BBT shift day estimated from the same record by the Sensiplan®
method.
Results Total agreement between the methods was obtained for 81% (294/364) of the
cycles and 18% (67) cycles differed by  1 day. For the 364 pairs of values distributed among
51 women the medians of the differences between the TTS transition day and Sensiplan®
initial day of the BBT rise (shift day) were not significantly different (C2  65.28, df  50,
p  0.07205).
Conclusion The advantages of the tracking signal algorithm are that in many cases it was
possible to identify the BBT shift day on that very day – rather than only some days later
– and to estimate the probability that a transition had occurred from the different values of
the tracking signal.

K E Y WO R D S Basal body temperature changes; Ovulation; Sensiplan®; Trigg’s tracking signal;

Symptothermal method
............................................................................................................................................................................................................

I N T RO D U C T I O N
in temperature is not completely understood. One
The rise in the basal body temperature (BBT) that explanation is the effect of progesterone on the tem-
occurs in a woman’s cycle is caused by the rising perature regulation centre in the brain4; another reason
progesterone levels1–3 after ovulation. Although the may be the effect of that hormone on the diameter of
temporal relationship between these two phenomena the subcutaneous vessels, leading to a stronger perfu-
has been well studied, the mechanism of the increase sion of the skin during the second part of the cycle5,6.

Correspondence: G. Freundl, Univ. Düsseldorf, Am Alten Rhein 9, D-40593 Düsseldorf, Germany. E-mail: freundlg@uni-duesseldorf.de

© 2014 The European Society of Contraception and Reproductive Health

DOI: 10.3109/13625187.2014.948612

DEJC_A_948612.indd 1 8/22/2014 2:07:13 PM

 Detection of significant basal body temperature changes
To indicate that ovulation has occurred and hence
that fertility is over for the current cycle, a significant
increase in the BBT must be found which is typically
of the order of at least 0.2oC on the third day of the
rise. Effective rules have been established to assist in
recognising this change7. The summary of the rules, in
connection with an evaluated system of teaching and
training, is called Sensiplan®8. In 2011 our teaching
and training organisation Malteser Arbeitsgruppe NFP
in Cologne registered the Natural Family Planning
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14
(NFP) method known as Sensiplan® with the Office
for Harmonization in the Internal Market (OHIM,
the official trademarks and designs registration office
of the European Union).
A possible alternative method for detecting the
BBT shift day was investigated based on the Trigg’s
tracking signal (TTS) algorithm9. This procedure was
developed to detect long-term changes in mean values
of signals used in control processes in industry. It spots
systematic increases or decreases that may indicate
deterioration of quality during manufacturing. Since
a BBT shift day is a day with a relatively large rise in
For personal use only.
BBT above previous baseline values, which is main-
tained for several days, it represents a departure from
the mean and hence may be detected in principle by
the TTS calculations. Previously, the TTS calculation
had been modified to detect successfully the first sta-
tistically significant rise in urinary excretion rates of
metabolites of ovarian hormones10 as markers for
delineating the limits of the fertile window. Therefore,
it was of interest to determine whether the TTS algo-
rithm might be adapted in a similar fashion and applied
in the analysis of BBT profiles as the cycle proceeds
day by day to detect the day of the BBT shift, and to
compare the results with those already obtained from
application of the Sensiplan® rules8. Identification of
the BBT shift day allows a woman to determine the
end of her fertile phase by using this information in
combination with other rules.
In summary, with this pilot study we wished to find
out whether the TTS transition day marks the crucial
rise of the BBT in women’s menstrual cycles as the
latter unfold. For this purpose we compared the results
of the Trigg’s signal algorithm with the results of Sen-
siplan®. If the agreement between the two should
prove satisfactory then it would be useful to perform
efficacy finding studies (EFSs) and effectiveness studies
(ESs)11,12 with different women using this tool. An EFS
is the first step in assessing the appropriateness of a
2 The European Journal of Contraception and Reproductive Health Care
DEJC_A_948612.indd 2
Freundl et al.
device for a woman who wants to avoid pregnancy.
An ES determines the effectiveness of the device in
preventing a pregnancy (user performance)12.
M AT E R I A L S A N D M E T H O D S
An Excel file from the German Natural Family
Planning database with daily BBT readings from 51
women aged 24 to 35 was studied. Of the 406 cycles
in the database only 364 could be used for the analysis
since 42 of the cycles had more than two consecutive
missing temperature values. The data for each cycle
were imported into a computer programme (NewDB)
written for this purpose and displayed as daily BBT
profiles available for analysis using the TTS algorithm.
The temperature charts were evaluated with respect
to a significant temperature rise using both Sensiplan®
and the TTS algorithm.
The Sensiplan® rules
In this method, the observations of the woman con-
cerning BBT, cervical mucus and other cycle symp-
toms are recorded using a standardised cycle chart13.
According to Sensiplan® a temperature rise has
occurred when there are three consecutive tempera-
ture readings, which are all higher than the preceding
six, with the third higher reading being at least 0.2oC
above the highest of the preceding six lower ones
(Figure 1a). The first of these three days with elevated
BBT values (rise days) is the ‘BBT shift day’. In prac-
tice there are two exceptions to this definition of the
BBT rise and identification of the BBT shift day:
(1) If the third temperature reading is less than 0.2oC
above the preceding six values, one must wait for a
fourth reading. The latter must also be higher than
the preceding six lower readings, but not necessar-
ily by 0.2oC (Figure 1b).
(2) Between the three higher readings there may be
one value, which drops down to, or below, the line
which covers the preceding six readings as shown
in Figure 1c. This lower value can be ignored.
However, the third higher reading must be at least
0.2oC above the cover line (Figure 1c). These two
exception rules cannot be combined7.
‘Disturbed’ temperature readings (‘Störungen’) are
excluded and are not to be taken into account: various
8/22/2014 2:07:14 PM
 Detection of significant basal body temperature changes Freundl et al.

‘disturbed’ temperature readings varies from woman to

woman.

An algorithm for detecting the day of

the shift in basal body temperature
(Trigg’s tracking signal transition day)

A significant change in BBT can be detected only

by comparing it with the BBT measurements from
preceding days. Given the variability inherent in BBT
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14

measurements, on day n of the cycle (or day n of a

continuous sequence of BBT measurements) the dif-
ference between the BBT value on that day and a
weighted average of the deviations (SFE) from the
preceding measurements (ESA) is compared with the
mean absolute deviation (MAD), which is a statisti-
cally robust measure of the variability. Trigg’s tracking
signal on day n is
SFEn
TTSn  , (1)
MADn
where
For personal use only.

« BBT3  ESA 2 n3

® (2)
SFEn  ¬

®­A BBTn  ESAn1  1 A SFEn1 n  4

and
Figure 1 Sensiplan® rules for evaluating the basal body
temperature (BBT) with the two exceptions to the rules «max ¨ BBT  ESA , 0.05·
® n3
for temperature rise. The average BBT ranges between ª 3 2 ¹ (3)
MADn  ¬
36.0 and 37.5oC. The abscissa shows the cycle days, and
®A BBTn  ESAn1  1 A MADn1 n  4
the ordinate, the BBT-values. (a) evaluation of temperature ­
(BBT)-rise; (b) first exception rule for temperature (BBT)-
evaluation; (c) second exception rule for temperature in which
(BBT)-evaluation.
ESAn  ¬

«®0.5 BBT  BBT
1 2 n2
(4)
­®ABBTn  1 A ESAn1 n  3
changes in the normal daily routine and conditions for
taking the temperature can disturb the morning read-
ing and sometimes cause a false shift. It cannot be
predicted which circumstances will affect an individual
woman’s temperature and which will not. She must
find this out for herself by making a note of anything
she believes may affect her waking temperature. Such
things can be: mistakes or variations in the measure-
ment method; measurements at different times; short
or disturbed nights; retiring late; unaccustomed
amounts of alcohol; emotional strain; excitement;
change of surroundings through journeys, holidays,
climate change, and more; changes in daily routine;
illness or indisposition. Therefore, the percentage of
The European Journal of Contraception and Reproductive Health Care
DEJC_A_948612.indd 3
and 0  A  1 is an empirical constant. ESAn, SFEn
and MADn represent different weighted averages of
both BBTn and of the earlier values (BBTn-m, m  1,
2, …, n – 1) used in the calculation of ESAn-1, SFEn-1
and MADn-1 (and their earlier equivalents). It is often
said that they are exponentially weighted because the
values for day n are based on the corresponding values
for day n – m (m  1, 2, …, n – 1) each of which is
multiplied by (1 – A)m, so earlier values contribute less
than more recent values. While SFEn incorporates the
direction of change of BBT (rise or fall), MADn does
not because it is a weighted average of the absolute
values of changes in BBT. As MADn must be greater
3
8/22/2014 2:07:15 PM
 Detection of significant basal body temperature changes
than zero, but SFEn does not, TTSn ranges from –1
to  1 and if either of these extremes is reached it is
statistically certain that a change has occurred on that
day. Other values of TTSn correspond to lower prob-
abilities that a change has occurred on that day (for
details of the method see Blackwell and Brown9). As
BBT rises near the time of ovulation, the value of
TTSn should change sign from a negative to a positive
value, the BBT value on that day must be greater than
each of the preceding six BBT values and TTSn must
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14
equal or exceed 0.61 (for A  0.3 this corresponds to
p ў 0.88 that a change has occurred) within three days
of that day (we refer to this day as the TTS transition
day).
For consistency with Sensiplan®, the search for the
TTS transition day started on the 6th cycle day, or the
6th day of the sequence of BBT values if the BBT
readings did not start on cycle day 1. This was also
convenient because it minimised any bias that might
have been introduced by the initialisation of TTS
based on the first two BBT measurements (Equations
2 and 3). If there was no clear transition day, a day
For personal use only.
corresponding to a local minimum in TTS was identi-
fied and the TTS transition day was taken to be the
first day on or after this which conformed to the
following supplementary rules:
(1) SFEn   0.05 on that day;
(2) BBTn – ESAn – 1  0.072 on that day; and
(3) ESAn  2  ESAn – 1  0.088, but if BBTn  1 or
BBTn  2 are missing this rule is advanced by one
day and if this is not satisfied by the second day
Freundl et al.
Statistical methods
The normality of the data was assessed by means of the
Shapiro-Wilk test and the non-parametric Kruskal-
Wallis test was used to ascertain that similarly dis-
tributed data were obtained from all individuals. The
probability indicated by a particular value of TTSn was
estimated from the results of Batty14 and Gardner15.
R E S U LT S
To compare the usefulness of Sensiplan® 7 with that of
the Trigg’s tracking signal in detecting the BBT-shift
day we examined 364 cycles. Of these, a full cycle
length was not recorded in 20 cases (in these the mini-
mum recorded length ranged from 22 to 41 days) and
the length of those for which a full length cycle was
recorded was 24 to 34 days (mode  27 days) in more
than 90% of cases (only 14 cycles were less than 24
days and 17 cycles were more than 34 days, Table 1).
These cycles were comparable with normal cycles of
adolescent and premenopausal women.
In the sample of 364 records, 294 (81%) of the
cycles were in total agreement and 67 (18%) cycles
differed by  1 day (Table 2). There were three (1%)
cycles where the two methods of evaluation differed
by  2 days or more. In one cycle, which has not been
Table 1 Length (days) of 364 menstrual cycles.
Cycle length, days Number of cycles Percentage
(n  2) then use ESAn  3  ESAn.  24 14 4
24 20 6
While these rules might seem intricate, they are easily 25 33 9
incorporated into the software. 26 42 12
The tracking signal algorithm was refined by com- 27 52 15
28 47 14
paring TTS transition days with the BBT shift day
29 45 13
determined using the Sensiplan® rules. Where there
30 37 11
was a discrepancy between the two data methods, the
31 16 5
computer algorithm was modified and any change that 32 11 3
increased the number of discordant pairs was rejected. 33 7 2
By careful iteration, it was decided that A  0.3. For 34 3 1
this value, the tracking signal transition day occurred  34 17 5
within one day of the BBT shift day identified using Total 344 100
Sensiplan® for over 90% of cycles. Based on the results
of Batty14 and Gardner15, TTS  0.61, 0.69 and 0.84 The full cycle length was not recorded on 20 records,
although the BBT shift day was. The minimum recorded
correspond to p
ў 0.90, 0.95 and 0.99, respectively, for length of these 20 cycles was 22 days and the maximum
A  0.3. was 41 days.

4 The European Journal of Contraception and Reproductive Health Care

DEJC_A_948612.indd 4 8/22/2014 2:07:16 PM

 Detection of significant basal body temperature changes Freundl et al.

Table 2 Number and percentage of women by difference between Sensiplan® and the Trigg’s tracking signal (TTS)
algorithm evaluation of the BBT shift day.

TTS transition day

both before and
TTS always the same day or a TTS always the same day or a after the Sensiplan®
Always the same daya day later than Sensiplan®b day earlier than Sensiplan®c shift dayd

23 17 5 6
45% 33% 10% 12%

aCycles
showing no differences between Sensiplan® and the algorithm (TTS).
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14

bWomen who had negative differences in one or more cycles but never a positive difference.
cWomen who had a positive difference in one or more cycles but never a negative difference.
dWomen who had both positive and negative differences.

included in the sample, missing values in the critical
period made computer analysis impossible although
the Sensiplan® rules coped with this example. For
37% of the cycles (109/294) where there was total
agreement the tracking signal algorithm was able to
identify the BBT shift on the day but for the remain-
der the supplementary rules were necessary to confirm
the BBT shift day.
For personal use only.
Sensiplan® days of the BBT rise were not significantly
different (C2  65.28, df  50, p  0.072) (Table 3).
Close examination of the data revealed that one
individual had consistently negative differences; after
removal of this woman’s records the medians of the
differences of the two methods in the remaining 98%
of the sample did not diverge significantly (C2  55.79,
df  49, p  0.234).

There was no significant difference between the The advantages of the tracking signal algorithm
BBT shift day determined with Sensiplan® and the are that in many cases it was possible to identify the
TTS algorithm (Table 3). According to the Shapiro- BBT shift day on that day, rather than some days later,
Wilk test, the differences between the two methods and to estimate the probability that a transition had
were not normally distributed (W  0.513, p  0.001),
so we used the Kruskal-Wallis test to assess the signifi-
cance of the observed differences. For 45% of women
the two methods consistently gave identical estimates
and in more than 90% of cycles the difference between
the methods was one day or less (Table 2). For the 364
pairs of values distributed among 51 women, the
medians of the differences between the TTS and
Table 3 Frequency of difference (Sensiplan® vs. Trigg’s
tracking signal [TTS] algorithm) in estimates of the cycle
day of the basal body temperature (BBT) shift in 364
menstrual cycle records.
Days differencea 2 1 0 1
Number of cycles 1 54 294 13
Percentage 0.3 14.8 80.8 3.6
aThe difference was calculated as the cycle day of the
BBT shift by Sensiplan® minus the corresponding cycle
day calculated from TTS. If the day of the BBT shift is
later than the day given by Sensiplan® the difference is
negative and vice versa. For these data the mean was
0.088  0.574 which was significantly different from 1
(p  0.0018).
The European Journal of Contraception and Reproductive Health Care
DEJC_A_948612.indd 5
occurred from the different values of the tracking
signal.
Examples of analysis of individual charts
Daily BBT values for one menstrual cycle are shown
in Figure 2A. Over the first 16 days there was a gen-
eral downward shift in the BBT values to reach a
minimum of 36.6oC on cycle day 16, before the ris-
ing values characteristic of a luteal phase shift are
seen starting on cycle day 18. The TTS decreased
in parallel over the same period to reach a value of
 0.981 on day 16. The transition between previously
negative TTS values to positive ones occurred on day
 2
18 and the TTS programme identified the TTS tran-
2 sition day also as occurring on day 18. This was also
0.5 the day that the woman recognised her mucus peak
symptom and, according to the rules of Sensiplan®, the
BBT shift day was also day 18. Thus, in this case, there
was complete agreement between the two algorithms
for the detection of the BBT shift day.
A second cycle is shown in Figure 2B. For this cycle
there were fluctuations in the BBT values during the
5
8/22/2014 2:07:16 PM
 Detection of significant basal body temperature changes Freundl et al.
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14
For personal use only.

Figure 2 BBT and tracking signal values for complete menstrual cycles. The BBT values are shown as solid circles
(‫ )׶‬and solid lines (‫)ד‬, and the tracking signal values as empty circles (‫ )ײ‬and broken lines (…). The horizontal
line shows the transition between negative and positive tracking signal values. The cycles came from four different
women (A–D).

follicular phase which gave a first transition of the TTS
(from negative to positive values) on day 6. However,
this was only the 5th BBT value in the sequence so
this transition was ignored according to the rules of
the algorithm and a second such day was sought by
the programme. The TTS transition day occurred on
cycle day 20 which was also the BBT shift day and
the subject’s mucus peak day was cycle day 18. Appli-
cation of the Sensiplan® rules gave a BBT shift day on
day 20, which was, again, in complete agreement.
An example in which the discrepancy in timing
between the two procedures amounted to one day is
shown in Figure 2C. In this example the subject’s
mucus peak day was cycle day 13 and the TTS transi-
tion day was day 12 which was also signaled as the day
of the BBT shift by the programme. The Sensiplan®
rules indicated cycle day 13 as the first day of the BBT
rise and hence the TTS transition day was one day
earlier (i.e., a positive difference of  1 day). An exam-
ple of a cycle for which the two methods differed by
two days is shown in Figure 2D. This cycle illustrates
the second exception to the rules of Sensiplan®.
There is a clear rise on day 12 signaled by the TTS
programme but the BBT value on day 13 has dropped
below the cover line. For this reason the Sensiplan®
rules indicate day 14 as the day for the start of luteal
phase infertility and the difference was  2 days.
DISCUSSION
Findings and interpretation
In every cycle there was a clear transition from nega-
tive- to positive TTS values (as in Figure 2), consistent

6 The European Journal of Contraception and Reproductive Health Care

DEJC_A_948612.indd 6 8/22/2014 2:07:16 PM

 Detection of significant basal body temperature changes
with the fact that the BBT values are biphasic in nor-
mal cycles. In about a quarter of the cases the TTS is
smaller than  0.9 by the end of the follicular phase
(Figures 2B and 2D) showing that there was a statisti-
cally significant low phase before the peri-ovulatory
rise. A TTS of this magnitude indicates virtual statisti-
cal certainty that the decline in BBT values through-
out the follicular phase is real. In principle, all that is
required to find the BBT shift day in these cases is
to note the day that the TTS changes from negative
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14
to positive. Consequently, in cases such as these the
tracking signal algorithm makes it possible to identify
this event on the day it happens rather than two or
three days later.
Strengths and weaknesses of the study
We have considered cycles from women aged 24–35
years for whom the detection of the fertile window
and achieving or preventing conception are especially
important. The cycles varied considerably in length
(from 22 to 46 days, Table 1) and most of them were
For personal use only.
complicated (Figure 2). Specifically, the BBT rose and
fell during the follicular phase, which obscured the
BBT-shift day. These fluctuations in BBT make algo-
rithms such as Sensiplan® necessary to accurately detect
the BBT shift day and the end of fertility. However,
Sensiplan® only identifies the BBT-shift day two or
three days after it has occurred (Figure 1). In con-
trast, the tracking signal algorithm often pinpointed
the TTS transition day on the day it occurred and,
even in the most complicated cases, it takes no longer
than Sensiplan®. Moreover, the two methods agreed
to within a day in 99% of the cycles we examined
and there was only one cycle from which the track-
ing signal algorithm could not estimate a TTS transi-
tion day (because of a lack of BBT measurements) but
Sensiplan® could.
The parameters of the tracking signal algorithm
were established by comparison with the Sensiplan®
results for this set of data. The algorithm now needs
to be tested against a larger set of data before its general
use can be advocated.
Differences in results and conclusions
in relation to other studies
The correlation between the Sensiplan® estimate and
ovulation and the fertile window has been studied for
The European Journal of Contraception and Reproductive Health Care
DEJC_A_948612.indd 7
Freundl et al.
a large number of cycles and is well established16–20.
The Sensiplan® methodology was primarily developed
to be used by the woman herself without any tech-
nical device. The tracking signal algorithm presented
here constitutes a new method that could be used in
conjunction with various devices such as computers,
phones or thermometers to determine the end of the
fertile phase in a prospective (i.e., as the cycle unfolds)
manner. To our knowledge such an approach in dating
the BBT rise in a woman´s cycle has not previously
been reported in the literature.
It should be emphasised that detection of the BBT
shift does not necessarily give a reliable end of fertility
marker as long experience with Sensiplan® shows.
It is necessary to incorporate other features of the
cycle such as the mucus peak day. Observing only the
temperature in a cycle gives a limited picture of fertil-
ity. However, by evaluating the temperature carefully
we can achieve a very reasonable effectiveness in pre-
venting a pregnancy by requiring sexual intercourse
to resume only on, or after, the third evening after the
morning of the BBT rise. This management is called
the ‘strict form of the temperature method’21,22.
Relevance of the findings: Implications
for clinicians
The detection of the first BBT-rise in a woman’s
cycle using the Trigg’s tracking signal transition day
could help a client to recognise the beginning of the
absolutely infertile phase in a cycle. In particular, the
temperature charting required for Sensiplan® might be
avoided. Moreover, such an approach might be attrac-
tive to women who look for technical assistance in
evaluating their daily BBT values.
Unanswered questions and future research
Three research questions should be tackled in the
future:
v Handling of temperature disturbances: the BBT may
be affected by disturbances like illness, measure-
ments at different times; short or disturbed nights,
stress, etc. Women who use the Sensiplan® method
learn to recognise their individual reaction to dis-
turbances and how to handle those disturbed values.
A further study should especially investigate the
results of the new algorithm in cycle charts with
severe or multiple temperature disturbances.
7
8/22/2014 2:07:16 PM
 Detection of significant basal body temperature changes
v Contraceptive efficacy: to assess the effectiveness
of such procedures to achieve, avoid, or delay a
pregnancy well-designed prospective effective-
ness studies (ESs) need to be carried out and
the algorithm validated on an independently
collected BBT dataset. In particular, positive dif-
ferences between the Sensiplan® method and the
TTS transition day mean that the TTS tran-
sition day occurs before the Sensiplan® BBT
shift day (as in Figures 2c and 2d) and hence
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14
may lead to unexpected pregnancies. While the
chances of this occurring are very low (0.6%,
Table 3), it is important to determine which signal
is the most ‘effective’ (reliable) in these cases.
v The BBT approach cannot give information on the
beginning of the fertile window of the cycle. However, it
may be of interest to study the periods when the
tracking signal is large and negative, and its relation-
ship to the LH peak and the mid-cycle oestrogen
peak. There is some evidence23–25 that high oestro-
gen values can have an inhibiting effect on the
basal body temperature. Hence a comparison of the
For personal use only.
statistically negative tracking signal values in rela-
tion to the rising oestrogen values at the time may
reveal a dependence that could be used in some
cycles to indicate the beginning of the fertile win-
dow. This should also be studied in relation to
REFERENCES
1. Döring G. Physiologie und Pathologie der Basaltemper-
atur bei der Frau und ihre diagnostische Bedeutung
[Physiology and pathology of the BBT of women and
its diagnostic meaning]. Fortschr Med 1965;83:885–7.
2. Barton M, Wiesner B. Waking temperature in relation
to female fecundity. Lancet 1945;I:663–8.
3. Fischer RH. Progesterone metabolism. III. Basal body
temperature as an index of progesterone production and
its relationship to urinary pregnanediol. Obstet Gynecol
1954;3:615–26.
4. Freundl G. Present stage of knowledge of reproductive biology
concerning natural family planning. GENUS LIV - n. 3–4,
57–74. Rome: Università degli studi di Roma ‘La
Sapienza’, Dipartimento di Scienze Demografiche
1998.
5. Gomez-Amor J, Martinez-Selva JM, Roman F, et al.
Electrodermal activity in menstrual cycle phases: A
comparison of within- and between-subjects designs.
Int J Psychophysiol 1990;9:39–47.
8 The European Journal of Contraception and Reproductive Health Care
DEJC_A_948612.indd 8
Freundl et al.
the cervical mucus characteristics prevalent at the
same time.
CONCLUSION
The advantages of the tracking signal algorithm we
describe are that in many cases it allows the identifi-
cation of the BBT shift day on that day – rather than
some days later – and to estimate the probability that
a transition has occurred from the different values of
TTS. Linking the daily BBT records to a web-based
device that incorporates the TTS algorithm modified
by the Sensiplan® vectors should ease the bookkeep-
ing strain placed on women to determine the shift
day, and inform those resorting to such a device of the
occurrence of the latter sooner and more reliably than
is possible with currently available means.
Funding: Only the maintenance of the German NFP-
database is funded by the German Catholic Bishop’s
Conference. This article reflects the views of the
authors only and is not influenced by this body.
Declaration of interest: The authors report no con-
flicts of interest. The authors alone are responsible for
the content and the writing of the paper.
6. Gomez-Amor J, Martinez-Selva JM, Roman F, et al.
Electrodermal activity, hormonal levels and subjective
experience during the menstrual cycle. Biol Psychol
1990;30:125–39.
7. Arbeitsgruppe NFP. Natürlich und sicher [Natural and
safe], 17th edn. Munich: Trias Verlag 2011.
8. Frank-Herrmann P, Sottong U, Baur S, et al. Natürliche
Familienplanung – Sensiplan – eine moderne, verläss-
liche Methode [Sensiplan – a modern and effective
method of natural family planning]. Gynäkologe
2011;44:17–22.
9. Blackwell LF, Brown JB. Application of time-series
analysis for the recognition of increases in urinary estro-
gens as markers for the beginning of the potentially
fertile period. Steroids 1992;57:554–62.
10. Blackwell LF, Brown JB, Cooke D. Definition of the
potentially fertile period from urinary steroid excretion
rates. Part II. A threshold value for pregnanediol
glucuronide as a marker for the end of the potentially
fertile period in the human menstrual cycle. Steroids
1998;63:5–13.
8/22/2014 2:07:16 PM
 Detection of significant basal body temperature changes
11. Freundl G, Godehardt E, Kern P, et al. Estimated
maximum failure rates of cycle monitors using daily
conception probabilities in the menstrual cycle. Hum
Reprod 2003;18:2628–33. 19.
12. Freundl G, Frank-Herrmann P, Gnoth C. Cycle monitors
and devices in natural family planning. J Reproduktionsmed
Endokrinol 2010;7[Special Issue 1]:90–6.
13. Arbeitsgruppe NFP, Malteser Köln. Accessed 16 June
2014 from: http://www.nfp-online.de/uploads/media/ 20.
Englisch_02.pdf
14. Batty, M. Monitoring an exponential smoothing
Eur J Contracept Reprod Health Care Downloaded from informahealthcare.com by Colorado State University on 09/04/14
forecasting system. Operational Res Q 1969;20:319–25. 21.
15. Gardner ES. Automatic monitoring of forecast errors.
Forecasting 1983;2:1–21. 22.
16. Gnoth C, Frank-Herrmann P, Bremme M, et al.
Wie korrelieren selbstbeobachtete zyklussymptome mit
der ovulation? [How do self-observed cycle symptoms
correlate with ovulation?]. Zentralbl Gynakol 1996; 23.
118:650–4.
17. Frank-Herrmann P, Gnoth C, Baur S, et al. Determina-
tion of the fertile window: Reproductive competence 24.
of women – European cycle databases. Gynecol Endocri-
nol 2005;20:305–12.
18. Jacobs C, Frank-Herrmann P, Jenetzky E, et al. 25.
For personal use only.
Self-observation of the fertile window in the menstrual
The European Journal of Contraception and Reproductive Health Care
DEJC_A_948612.indd 9
Freundl et al.
cycle by the infertility patient: diagnostic relevance
and spontaneous pregnancy rate. Hum Reprod 2006;
21(Suppl. 1):i194.
Frank-Herrmann P, Heil J, Gnoth C, et al. The effective-
ness of a fertility awareness based method to avoid preg-
nancy in relation to a couple’s sexual behaviour during
the fertile time: A prospective longitudinal study. Hum
Reprod 2007;22:1310–9.
Raith-Paula E, Frank-Herrmann P, Freundl G, et al.
Natürliche Familienplanung heute [Natural family planning
today], 5th edn. Heidelberg: Springer Verlag 2013.
Döring G. Die temperaturmethode [The temperature
method]. Stuttgart: Georg Thieme Verlag 1982.
Döring G. Über die zuverlässigkeit der Temperaturmeth-
ode zur Empfängnisverhütung [On the reliability of the
temperature method for contraception]. Dtsch Med
Wochenschr 1967;92:1055–61.
Davis M, Fugo N. The cause of physiologic basal body
temperature changes in women. J Clin Endocrinol
1948;8:550–63.
Magallon DT, Masters WH. Basal temperature studies
in the aged female: Influence of estrogen, progesterone,
and androgen. J Clin Endocrinol Metab 1950;10:511–8.
Stephenson LA, Kolka MA. Thermoregulation in
women. Exerc Sport Sci Rev 1993;21:231–62.
9
8/22/2014 2:07:16 PM