Estimation for the prevalence of syphilis in Japan (Theory of Biomathematics and Its Applications XII : Mathematical and experimental approach to clarify patterns in a transition process)

Title:

Estimation

for

the prevalence ofsyphilis in

Japan

Yuichiro

Miyamatsul

and Hiroshi

Nishiura1

$|$

Graduate School ofMedicine,The University ofTokyo,[email protected]

Keywords:syphilis, MSM, natural history

Introduction:

Overthe pastdecades. thenotified number ofsyphilis patients iniapan had steadily

decreased,perhapsowingto $pro\searrow\fbox{Error::0x0000}$ision ofantibiotics and improvementin the

manner

ofsexual

intercourse

among

prostiUttes. Nevertheless. there has been

a

recent increase in syphilis

nolificationsespecial

among

MSM $\{mcn$having

sex

with men) in $()rl\backslash a\iota\tau$

areas.

$Col\cdot\iota$sidering

that the natural history of syphilis involves long incubatioR period and asymptomatic

infection,all infected int$i$\vee$iduals

are

notcounted bymedical attendance.$3^{{\}}Ise\iota$₎$\iota\cdot e\dot{s}et\backslash tec4$ study

aimed to statisticallyestimatetheincidcnce and prevalence ofsyphilis

over

timcand age in

Japan.

Methods:

Amathematical model that describes the natural history ofTreponema pallidum

was

devised

to fit the observablevariables toempiricaldata andestimate undiagnosed fractions ofcases.

Estimatingthetime dependentgrowth of infected individuals ineach disease-stageand jointly quantifyingthe time-dependent rateof diagnosis, the incidenceas well

as

$P^{levalence}$

of syphilis

are

estimated.Amaximum likelihood method

was

applied. ushgthe datasets of

notified $sy$_})hilis

cases

from the National Surxei lance Report.

Results:

InFigure

1.

which isthe comparison between observed and expecteddata,

we were

ableto

obtain the goodness of fit in all stages, despite ofthe factthat primaryand secondarysyphilis

were

combined in the sulveillance data inthe recent $yea\iota\cdot s$after200]. Figure2 shows the

yearly incidence

as

newlyinfected

incidences.

lambda. Inthe recentyears,

we were

able to

estimatethe increase inmale’s incidence.which

can

explainthe recentresurgence. In

addition.

atthattime,_{the incidence in female started to} $inc\iota\cdot$

ease.

Figure 3 illustratesthe rates

ofdiagnosisina114stages.These diagnosticrates increased in $199(\rangle_{3}$

.

mostlikely duetothe

AIDS panic, contribntingpeople to testtheirSTIs

more

frequently in clinics. Inthe recent5

数理解析研究所講究録

years,

diagnosticrates

are

quite low in all stages.

Conclusion:

Wehave identified thatdramatic decline in incidence

over

50 years from

over

5000

individualsper yearto below

1000

individuals in recentdecades. There has been therecent

resurgence.

Themost recent incidence

was

about

4000 persons per year.

compal.ableto late

$i960s$. _Resur

ence

was

_notcaused by increasedascertainment in recentyears

Figure 1. Comparison between observed and expecteddata

Figure2. Yearly incidence (lambda)

$\simarrow\prime\overline{s}^{n\emptyset\phi a\prime r\wedge\infty\wedge\sim}\aleph^{\omega’\grave{w}\overline{\infty}\alpha\infty}\underline{\Phi}_{\wedgearrow\simarrow\sim\cdot\sim}Q{\}^{m\wedge\approx_{\underline{\tilde{o}}\underline{\alpha}\alpha}}\underline{{\}}$${\}\overline{\S}$

\S

$\not\in\triangleright$

\S $\S_{\overline{\tilde{S}}\tilde{8}}^{\kappa}$

Figure3.?herate ofdiagnosis

Primary syphilis Seoondarysyphilis

Latent syphilis Tert$ary syphilis

Contact:Yuichiro Miyamatsu

Graduate School ofMedicine. $/I^{\neg}he$University $ofTok\rangle$

$7arrow 3arrow t$ Hongo, Bunkyo-ku. Tokyo, 1$13-\zeta t03$

.

Japan

Tel:$+813$ 58413688_Fax:$+813$ 58413637

email:_{$yy.\cdot C^{2}$}