By Carol A. Kemper, MD, FACP

Clinical Associate Professor of Medicine, Stanford University, Division of Infectious Diseases, Santa Clara Valley Medical Center

Dr. Kemper reports no financial relationships relevant to this field of study.

Complications of Typhoid Fever

SOURCE: Cruz Espinoza LM, McCreedy E, Holm M, et al. Occurrence of typhoid fever complications and their relation to duration of illness preceding hospitalization: A systemic literature review and meta-analysis. Clin Infect Dis 2019;69(Suppl 6):S435-S448.

These authors performed a meta-analysis of published reports of typhoid fever treatment to determine whether complications occur more frequently in children vs. adults, and the degree to which delay in diagnosis and treatment increases the risk for complications. Only studies with culture-confirmed disease were included in the analysis (stool, blood, bone marrow). The authors reviewed eight published studies from seven different countries that included sufficient descriptive data for use in meta-analysis. Six of these included persons of all ages, five focused only on treatment in adults, and two only on treatment in children. Originally it was believed that patients presenting more than two to four weeks into their illness would be at increased risk for complications. However, examination of pooled data found that patients presenting 10 days or more after symptom onset were at increased risk for severe disease, and this was used as a cut point for data analysis.

The pooled prevalence of complications in hospitalized patients with positive blood cultures for Salmonella typhi was 27% (95% confidence interval [CI], 21-32%), although there was significant variability in the frequency of reported complications between the 13 studies. Examination of pooled data revealed that patients presenting 10 days after the onset of symptoms were at greater risk for complications and hospitalization compared with those presenting with < 10 days of symptoms (36% vs. 16%), and were at three times greater risk of severe disease (odds ratio [OR], 3.0; P < 0.0001). Pooled prevalence data also showed that complications in children were significantly higher than in adults (27% vs. 17%), although, again, there was significant variability between the studies.

The most frequent complications observed in pooled data were encephalopathy (7.3%), gastrointestinal bleeding (5.2%), and nephritis (4.8%). However, the most frequent complications in those presenting for care < 10 days after symptom onset were hepatitis (5.1%) and gastrointestinal bleeding (4.0%). In contrast, complications in those presenting 10 days into their course included encephalopathy (18%) and gastrointestinal bleeding (14%). Less frequent complications, occurring 10 days after symptom onset, and which were unusual in those presenting earlier in their course, included myocarditis (3.9%), cholecystitis (3.4%), osteomyelitis (2.2%), and intestinal perforation (1.1%).

Complications of typhoid fever in bacteremic patients requiring hospitalization are much more frequent than previously estimated, based on pooled prevalence data. Risks for more severe disease are much greater in children than adults, and in those presenting with 10 or more days of symptoms before diagnosis and treatment is begun.

Do I Smell a Rat? Or Is the Rat Smelling Me?

SOURCE: Fiebig L, Beyene N, Burny R, et al. From pests to tests: Training rats to diagnose tuberculosis. Eur Respir J 2020;55:1902243.

Researchers have spent years training giant Gambian rats to sniff out tuberculosis (TB) in sub-Saharan Africa. Preliminary data from this work were presented in this column in 2003 and 2010. By 2010, researchers reported that trained technicians detected positive acid-fast bacilli (AFB) smears by light microscopy in 10,523 specimens, but the rats performed slightly better, detecting an additional 600 specimens missed by the technicians. Overall, Gambian rat sniffing was 89% sensitive for detection of TB in expectorated specimens, compared with culture results.

Since then, the rats have continued to improve, detecting more than 16,000 cases of TB in three different African countries, with a sensitivity of > 90% and a specificity of > 70%. This is better than World Health Organization requirements for laboratory methods for detection of TB. But there was one major hurdle: getting the “rat test kit” to the clinical site. Previously, the rats were sniffing heat inactivated samples at a central research laboratory, which, since 2018, was performing confirmatory molecular testing also. Significant delays were occurring in getting specimens to the lab and the results back to the communities by which time many patients were lost to follow-up. By moving the lab closer to the communities being tested, and using a network of motorbike couriers, samples have been processed more quickly. (No, the rats did not get to ride the motorbikes.)

But an interesting question remained. What were the rats actually smelling in the specimens? Analytical chemistry provided a surprising answer. Like many active infections, different bacteria exude a pattern of specific volatile compounds. It turned out that the rats were indeed smelling TB infection. While TB may share certain volatile chemicals with other bacteria or non-tuberculous bacteria, it was a complex pattern of volatiles specific to TB that the rats were detecting. It is odor pattern recognition much like waking up in the morning and smelling the coffee.

Cloth Masks — Just for Looks?

SOURCE: MacIntyre CR, Seale H, Dung TC, et al. A cluster randomised trial of cloth masks compared with medical masks in healthcare workers. BMJ Open 2015;5:e006577.

Friends and family are busy sewing charming face masks with sunflowers and kitties. My sister had not had her sewing machine out for 20 years until forced to shelter at home with little to do, and made beautiful masks for her family.

But how effective are these fabric masks? This randomized clinical trial evaluated the effectiveness of cloth vs. regular procedural masks in 1,607 healthcare workers (HCWs), recruited for study at 14 different hospitals in Hanoi, Vietnam, in 2011. The HCWs worked full-time in 74 high-risk units of the hospital, including the emergency departments, intensive care, and infectious disease units. The HCWs were randomized to wear a regular procedural mask, cloth mask, or no mask throughout an eight-hour shift for four weeks. Then they were followed for an additional one week for signs of respiratory illness. Either two procedural masks were provided to each worker per shift, or five cloth masks were provided per month, which were to be washed and rotated throughout the four-week study period. HCWs also were asked to wear their masks throughout their shifts, except for tea, meal, and bathroom breaks. Each HCW maintained a diary of the number of patient contacts per shift and their activities, including suctioning, sputum induction, intubation, bronchoscopy, etc. At the first onset of symptoms, HCWs reported for evaluation and respiratory polymerase chain reaction (PCR) panel testing.

HCWs had an average of 36 patient contacts per day. Those with cloth masks reported that they washed their masks 23 of 25 days (97%), either at home (80%), using the hospital laundry (4%), or both (16%).

Despite the fact that workers reported better compliance with cloth masks, an intent to treat analysis showed that rates of clinical respiratory illness, influenza-like illness (ILI), and laboratory-confirmed respiratory infection were significantly higher in the cloth mask group compared with the procedural mask group. Laboratory confirmation of respiratory viral infection was detected in 31/659 (5.5%) of cloth mask users vs. 19/580 (3.3%) of the medical mask users. The relative risk of ILI in the cloth mask group, compared with the other groups, was 13.25. Surprisingly, no significant difference was observed in the rates of infection between those wearing medical masks and those without masks. The reported frequency of hand washing was found to be significantly protective against clinical respiratory illness.

In the laboratory, the penetration of particles through cloth masks was significantly higher (97%) than with either medical masks (44%) or N95s (< 0.01%).

In addition to barrier protection, it is conceivable that other factors may be responsible for this observed difference in the risk of respiratory illness. Workers may re-adjust cloth masks more often than medical masks. Certain types of cloth used for cloth masks may be better at “acquiring” viral particles as workers move through their day, or the warm moisture from breathing may allow improved virus survival on masks.