[Table of Contents]

Volume: 24S5 - September 1998

Canadian Integrated Surveillance Report for 1995 on Salmonella,
Campylobacter and Pathogenic Escherichia coli

Human Salmonellosis Cases

There were 7,307 human cases of salmonellosis reported through provincial laboratories to the National Laboratory for Bacteriology and Enteric Pathogens (NLBEP) database. This was slightly lower than the numbers reported in 1993 and 1994, of 8,057 and 7,441 respectively. Summary data for 6,389 cases were reported and verified as correct by provincial and territorial public health units to the National Notifiable Diseases (NND) database. Rates for the1995 population (Statistics Canada) from the NLBEP and the NND Summary (NNDS) databases, by province, are presented in Figure 1. NND individual case data were received from seven provinces and one territory, for 6,061 cases. The differences in number of cases between these databases likely reflect the different chains of reporting and inherent gaps within each path. Different reporting paths among provinces will also affect the number and type of cases reported. Therefore, variation in reported rates of salmonellosis among provinces, as presented in the Figure, must be interpreted with caution.

Figure 1
Rates of human cases of salmonellosis (per 100,000 population) as reported through the National Notifiable Diseases system (upper numbers) and National Laboratory for Bacteriology and Enteric Pathogens (NLBEP) system (lower numbers) by province for 1995

Top 10 Serovars

The top 10 serovars from human cases reported to the NLBEP database in 1995 are listed in Table 1. Cases involving these serovars represented 72% of the cases for which the serovar was specified (62% of all cases).

Table 1 Top 10 Salmonella serovars from human cases reported in Canada, 1995
	Serovar	Cases '95 ('94)
1	S. typhimurium	1,366 (1,385)
2	S. enteritidis	964 (1,255)
3	S. heidelberg	670 (507)
4	S. hadar	597 (741)
5	S. thompson	286 (220)
6	S. agona	166 (122)
7	S. newport	127 (142)
8	S. typhi	120 (136)
9	S. infantis	114 (121)
10	S. saintpaul	104 (127)

The top five serovars were unchanged from 1994, except that S. heidelberg and S. hadar exchanged positions at 3 and 4. S. infantis replaced S. berta in the top 10 serovars.

New and Emerging Serovars

Three serovars, S. istanbul (7 cases), S. tilene (6 cases), and S. wassenaar (6 cases),were associated with more than five cases in 1995 and had not been identified in 1993 or 1994. S. barielly and S. kiambu, which were similarly identified as new in 1994, were reported again in 1995 at similar levels to those seen in 1994.

Table 2 shows five serovars that were associated with 20 or more cases in 1995 and that demonstrated a substantial increase over the 1994 values. ¹
None of the serovars identified as emerging in 1994 is listed again in this group in 1995. There may be considerable inherent fluctuations in counts of individual serovars from year to year. Therefore, confirmation of  the emergence of some serovars will require evaluation of long-term trends.

Table 2 Serovars attributed to an increased number of cases, 1995
Serovar	1995	1994	% increase	Comments
4,5,12:b:	54	22	145%	No outbreaks reported
S. litchfield	25	10	150%	No outbreaks reported
S. norwich	21	3	600%	No outbreaks reported
S. panama	30	7	329%	No outbreaks reported
S. stanley	92	29	217%	No outbreaks reported

Long-Term Trends The total annual number of Salmonella cases reported from the NLBEP as well as of cases associated specifically with S. typhimurium decreased from 1985 to 1995 while the annual number of cases associated with S. enteritidis increased (Figures 2 and 3).

The pattern of Salmonella cases reported from the NNDS database was similar to that from the NLBEP database.

There were no consistent patterns among reported cases for the individual top 10 serovars over this 10-year period (Figures 2-6: note scale changes among Figures). S. typhimurium and S. infantis showed decreasing annual counts. Patterns of increasing and then decreasing counts are seen for S. hadar and S. agona, while the remaining six serovars demonstrate relatively uniform patterns of annual incidence.

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

1995 Trends

All serovars combined showed a seasonal trend, with increased rates of salmonellosis in the summertime (Figure 7). However, the distribution of cases over time varied greatly by serovar and province (Figures 8-11). For instance, as in 1994, S. enteritidis cases were reported predominantly from Ontario westward (Figure 9). Again in 1995, there was a higher rate of infection with S. heidelberg in eastern Canada than the rest of the country (Figure 10). S. heidelberg was the most common serovar of Salmonella reported from New Brunswick, Nova Scotia and Prince Edward Island in 1995. The cases occurred throughout the year, but most occurred from August through December.

Figure 7
All cases of Salmonella 1995 - Rate per 100,000 population* * per 200,000 for PEI Data from NLBEP

Figure 8
S. typhimurium cases 1995 - Rate per 100,000 population* * per 200,000 for PEI Data from NLBEP

Figure 9S. enteritidis cases 1995 - Rate per 100,000 population* * per 200,000 for PEI Data from NLBEP

Figure 10S. heidelberg cases 1995 - Rate per 100,000 population* * per 200,000 for PEI Data from NLBEP

Figure 11 S. hadar cases 1995 - Rate per 100,000 population* * per 200,000 for PEI Data from NLBEP

Non-fecal Source Isolates

Salmonella was isolated from feces in 4,739 cases, blood in 149 cases, urine in 103 cases and other non-stool sources (abscess, abdomen, wound, etc.) in 24 cases. The most common serovars isolated from blood were S. typhi (52), S. heidelberg (25), S. enteritidis (19), S. paratyphi A (13) and S. typhimurium (12). Children < 1 year of age accounted for 12% (13) of the cases with bloodborne infections. The variety of serovars isolated from urine was much greater than from blood. The most common serovars isolated from urine were  S. heidelberg (16), S. hadar (15), S. typhimurium (10) and S. thompson (8). Females accounted for 75% of the urine-isolated cases, suggesting that these are primarily ascending urinary tract infections.  Children < 2 years old represented 11% of urine-isolated cases and 14% of blood-isolated cases (compared with 24% for all cases, stool and non-stool).

S. typhi and S. paratyphi Cases

There were 120 cases of S. typhi reported in 1995 in the NLBEP database (only nine in the NND databases, but 45% of cases did not have serovar specified) as well as 23 cases of S. paratyphi A  and 19 cases of S. paratyphi B. The S. typhi isolates were from stool (42%) and blood (52%). Travel history was indicated for five cases. Only 9% of S. typhi and S. paratyphi cases occurred in children < 2 years old.

Hospitalizations and Deaths

From the NND Individual Case database, there were 244 (227.6/1,000 cases) inpatient and 38 (35.4/1,000 cases) outpatient hospital visits associated with Salmonella cases in 1995. Ten deaths (34.0/1,000 cases) associated with Salmonella were reported. Because hospitalization and outcome information was reported in less than 20% of Salmonella cases, only those cases for which information was provided were used to calculate these rates, which therefore may be biased. HIV status was not provided, so the impact of AIDS on hospitalizations and deaths could not be evaluated.

Outbreaks

There were 184 cases associated with 52 outbreaks reported in the NLBEP database for 1995 (Table 3). This is similar to the number of cases reported from outbreaks in 1994. Most of the outbreaks were reported to be family (44), community (5), restaurant (2) or nursing home (2) associated. The largest number of outbreaks was associated with S. typhimurium (14), followed by S. enteritidis (11), S. heidelberg  (6), S. hadar (5), S. montevideo (2 ) and S. thompson (2). A complete list is provided in Table 3. Food was implicated in three outbreaks (one salmon, one turkey and one unspecified), and animals were implicated in one outbreak (iguana).

In the 1995 NND database, 182 cases were associated with 40 outbreaks, and these were reported from only Ontario (161), Saskatchewan (12) and British Columbia (9). There is no indicator to link cases with specific outbreaks, but using date, serovar and the first three digits of the postal code, the cases could be grouped into apparent outbreaks. Thus, 40 outbreaks were reported to be associated with the following serovars: S. enteritidis (8 outbreaks; 81 cases),  S. typhimurium (5; 10),  S. thompson (3; 8), S. agona (2; 2), S. infantis (1; 6), S. saintpaul (1; 4), S. hadar (1; 3), S. brandenburg (1; 3), S. heidelberg (1; 2), S. branderup (1; 1) and Salmonella sp (unspecified) (16; 62). Only five of the 40 outbreaks appeared to match one of the 23 outbreaks identified in the NLBEP database for these provinces. This illustrates the underreporting of outbreaks in these databases.

A risk factor variable was completed for 12% of the cases in the NND database. Of these, 14% (106) had reported travel, 14% (107) had reported home and 8% (58) had reported restaurant as a risk factor for the case. Other reported risk factors included day care 1% (5) and workplace 1% (5).

Selected Phage Types

Summaries of the proportion of isolates that were phage typed and number of isolates of the major phage types for the top four Salmonella serovars are presented in Tables 4 and 5 respectively. These data are from NLBEP and HAL - Guelph databases. The different patterns of distribution across time and provinces, as well as the different non-human sources, may be partially due to variation in assessment and reporting processes or may indicate that these phage types have different host specificity, modes of transmission and pathogenicity. An isolate is much more likely to be assessed for phage type if it is associated with an outbreak. Table 4 shows the variability from year to year of phage typing among serovars and between human and non-human isolates of the same serovar.

Isolates of S. typhimurium PT 104 (DT104) were identified in 48 human cases and from two non-human sources. The number of cases appears to be increasing (10 human 1994; 3 human 1993), but the number of non-human isolates does not show a consistent temporal pattern (20 in 1994; 4 in 1993). However, approximately twice as many human S. typhimurium isolates were phage typed in 1995 than in 1994, whereas fewer non-human isolates were phage typed (Table 4).

In 1995, most of the human S. typhimurium PT 104 isolates were from British Columbia (42: 1.1/100,000); five isolates were from Quebec and one was from Saskatchewan. The cases were observed between the months of June and December, 40% (19/48) of them occurring in October. No age or gender association was observed. The non-human isolates were obtained from cattle in British Columbia.

Details for the other phage types are provided in Appendix B.

Age Distribution

The age distribution of cases reported to the NND Individual Cases database for each of these three pathogens showed a peak in children < 2 years of age, which extended to about 10 years of age (Figure 36). Campylobacter and Salmonella cases showed smaller secondary peaks in the 20-35-year-old age range, but no such peak was observed for E. coli cases.

Figure 36
Number of cases by age in 1995

The lower number of verotoxigenic E. coli and Campylobacter cases in the < 1-year-old children compared with 1-year olds is at least partially due to the fact that babies born in 1995 were at risk for only half the year, on average. This suggests that risk of salmonellosis is much higher for children < 1-year old than for 1-year olds.

The proportion of cases < 1 year of age was higher for Salmonella (8%) than for E. coli (4%) and Campylobacter (2%). However, the proportion of cases < 10 years of age was higher for E. coli (40%) than for Salmonella (32%) and Campylobacter (20%).

¹ Emerging serovars were associated with more than 20 isolates in 1995 and had counts such that the difference of the log of the 1995 count and log of the 1994 count exceeded the 75% upper confidence limit of a distribution of the difference for all serovars.

[Previous] [Table of Contents] [Next]