The new syphilis notification system in Germany
The new syphilis notification system, introduced in 2001, is operated and maintained by the Robert Koch-Institute (RKI) and is a passive, anonymous reporting system. All laboratories are required to report each positive syphilis laboratory result within two weeks using forms provided by the RKI. Laboratories are advised not to report clearly identifiable follow-up tests of adequately treated patients. The form consists of one original page and two copies bearing the same identification number, with instructions on how to complete them. The reporting laboratory completes the original page and posts this page directly to the RKI (reply postage paid). One copy remains with the lab to facilitate necessary clarifications. One copy is sent, in conjunction with test results, to the physician who completes the epidemiological and clinical section and sends the completed reporting form directly to the RKI. Identifying parameters for the patient, required for the anonymous reporting, are gender, month and year of birth and the first three digits of the five-digit postal code. If the postal code of the patient is not provided, the code of the physician or the laboratory is taken as surrogate.
Laboratory and clinical parameters have been defined, which are required for a report to fulfil the case definition [BOX and TABLE 1]. Inconsistent and missing information is checked individually by phone as far as possible.
One critical aspect of quality control of the data is checking for double (or multiple) reports, which is aided by an automatic search tool of the database. Upon entering a new report, this search feature produces a list of reports with the same sex and birth date (month/year), containing several additional key parameters. Multiple reports of the same event can thus be excluded with high reliability.

In addition to syphilis notification, a sentinel system for STIs was established in November 2002. Sentinel sites, which include private practices, hospital-based STI clinics and local health authorities, report cases of syphilis and other STIs to the Robert Koch-Institut. Patients are asked to provide information on sexual behaviour and social status on a self-administered questionnaire. Details of the methods are reported elsewhere [3]. In 2003, 311 cases of syphilis (11% of the notifications) were reported in parallel by the sentinel surveillance system.

Development of the syphilis epidemic in Germany
Time trends
In the final years of the previous reporting system, about 1100-1150 cases of syphilis (1.3-1.4/100 000) were reported each year in Germany [FIGURE 1]. Substantial underreporting was assumed; the proportion of unreported cases was estimated at about 30-40% of reported cases [4]. Reporting from the private health care sector and syphilis diagnoses in MSM were probably under-represented in the previous system, as indicated by the abrupt increase of cases from larger cities after introduction of the new reporting system, while syphilis incidence in women has remained stable at low levels [FIGURE 2]. On the other hand, since there was no case definition and limited quality control of reports before 2001, there may have been over-reporting of serologically reactive, but clinically inactive forms of syphilis, and there may have been double reporting mainly due to referrals between the private and the public sector.

The reported incidence of syphilis in Germany had been decreasing since the late 1970s and stayed stable at a low level throughout most of the 1990s until 2001 [FIGURE 1]. At the end of the 1980s the number of cases fell notably among men, probably as a result of changed behaviour in response to the emerging HIV/AIDS epidemic.
Outbreaks of syphilis were observed in Hamburg since 1997 and in Berlin since 1999. These were outbreaks among MSM, with most cases in the 30-40 year age group. According to a local study [5], a high percentage of cases (80%) in the Hamburg outbreak during 1997-98 occurred among HIV positive MSM. The Hamburg outbreak was followed by an increase of syphilis cases among men in Berlin in 1999 - it should be noted that since 2000, in the greater Frankfurt region, the increase of cases was not yet reflected by the surveillance system at that time, but was suggested by reports from dermato-venerological practices with mainly MSM patients and in Berlin it was supported by a shift in the male-to-female ratio of reported cases. The increase affected Cologne and some cities in the Ruhr region from 2000-2001, and Munich as well as other cities in Bavaria from early 2002 [6].
The number of reports fulfilling the case definition increased continuously from 1687 in 2001 to
2422 in 2002 and to 2932 in 2003 [7]. The completeness of reporting has been checked and was above 95% since the introduction of the new reporting system, thus only a negligible part of the increase can be attributed to an increase of the number of laboratories that contributed reports.
The clinical stage at diagnosis has remained relatively stable since 2001. The proportion with missing information decreased, while the proportion of latent syphilis increased [FIGURE 3].

Geographic and demographic aspects and affected populations
Syphilis cases, especially cases among MSM, are clustered in larger cities. The cities with the highest incidence rates are Frankfurt, Cologne, Berlin, Hamburg and Munich. The geographical pattern of incidence rates in 2003 by postal code areas is shown in figure 4.

While syphilis incidence in women remained stable (0.68 per 100 000 population in 2001, 0.65 in 2003), the proportion of cases diagnosed in women decreased from 15.5% (2001) to 9.4% (2003). Accordingly, syphilis incidence in men increased from 3.3 per 100 000 population in 2001 to 6.5 in 2003. The incidence among males peaks in the age group 30-39 years (17.1/100 000), while among females the incidence peak has shifted from the age group 25-29 years in 2001 to the age group 20-24 in 2003 (2.4/100 000).

Before January 2001, with the previous notification system, no information was collected on the probable route of infection. Information on probable route of infection was available for 66% of the notifications made during 2003, up from 57% in 2001. The most frequently reported route of transmission was sexual contact with other men (76% in 2003, up from 61% in 2001 and 70% in 2002). If we assume that the cases with unknown risk have a similar distribution to those with a known route of transmission, it can be estimated that currently around 75% of all syphilis cases notified in Germany are related to sexual contact between men. This finding is supported by similar observations in the sentinel surveillance system. Heterosexual contact is reported as infection risk in 23% of notifications with risk information. In the years 2001–2003, 23 cases of congenital syphilis in newborns were reported. Most of these children were born to mothers originating from countries other than Germany, which resulted in limited or delayed access to pre-natal care. In some cases, a first screening test in early pregnancy was negative and infection occurred during pregnancy.
Compared with the general population, a disproportionately high share of women with syphilis, and of patients with heterosexual intercourse as reported route of transmission, originate from central and eastern European countries [TABLE 2].

In self-defined homo- and bisexual men, who make up about 3-4% of the adult male population [7,8], syphilis incidence is much higher than in the rest of the population. In the most heavily affected group of homosexual men between 30 and 39 years old, the nation-wide incidence of syphilis is estimated to be about 100 cases/100 000. In metropolitan areas, the incidence of syphilis is up to seven times higher [FIGURE 2], but also the proportion of MSM in the population is probably about double that in towns and villages. Among HIV-positive MSM, who are disproportionally over-represented among MSM with syphilis (between 40-50% according to sentinel surveillance data), incidence rates above 1000/ 100 000 have been reached. In a recently conducted sexual behaviour survey among MSM, which probably oversamples HIV-positive men, 8% of the participating HIV-positive men reported a syphilis diagnosis in 2002 [9].

Discussion
The new laboratory-based reporting system for syphilis was introduced in Germany at a time of successive outbreaks of syphilis among MSM. Because of this coincidence, and because underreporting in the previous system was expected especially from the private sector where most MSM are diagnosed and treated, a reliable estimate of former underreporting rates with the physician-based reporting system is not possible.

Since the increase of reported cases of syphilis coincided with the implementation of the new reporting system, it was necessary to investigate whether the increase reflected an actual rise in the number of syphilis cases or resulted from the change of the reporting procedure. Both factors seem to play a role. Since the implementation of the new system, the notifications of syphilis have not increased in all regions, but mainly in metropolitan areas. The increase was less abrupt in Hamburg compared to other large cities, probably because the local outbreak investigation [5] had led to improved reporting compared to other cities. The increase has been continuous since the introduction of the new reporting system, with only a very slight increase in the number of participating laboratories. Private practices and STI clinics in affected areas report a significant increase in syphilis infections especially in MSM since about the year 2000, while the incidence rates for women before and after 2001 are similar. The steep increases in syphilis incidence rates among MSM in large cities are in line with the trend seen in other European countries and in North America [10,11].

The current syphilis epidemic in Germany, as reflected by the new reporting system, is characterised by successive outbreaks of syphilis among MSM in all larger cities in Germany, resulting in a sustained increase in syphilis incidence levels in this population group. The absolute number of heterosexually transmitted cases of syphilis in Germany seems to be relatively stable, though the proportion of heterosexual cases is decreasing. However, the relatively high number of heterosexually transmitted cases seems larger than in many other western European countries, probably reflecting the impact of population movement between syphilis high incidence regions such as eastern and south-eastern Europe and Germany.

Increasing awareness of the re-emergence of syphilis in Germany, as reflected by high media coverage, syphilis-awareness advertisements in the gay press and increased distribution of written information materials on STIs to MSM, has so far been insufficient to curb the spread of syphilis among MSM. Other control measures like increased offers for screening sexually active MSM have been recommended in a common statement of the RKI, the German STD society and the German AIDS society. Introduction of such measures as well as an increase of low threshold STI screening and treatment facilities was also discussed between the RKI, self-help organisations of gay men and local health offices in larger cities. However, the implementation of these measures is severely hampered by efforts to reduce health care spending (formally not allowing routine screening procedures paid by health insurance except in pregnant women; introduction of a consultation fee of € 10 per quarter year for every consultation with a physician) and reductions in public investment in public health (i.e. budget reductions and reduced staff for local health offices, resulting in restriction of STI services instead of expansion).