The emergence of RT027/ST1 contributed to an initial drastic increase in the incidence and severity of CDI in North America. In 2011, the annual national burden of CDI in the United States reached an estimated 476,400 cases, based on data extrapolated from the US Centers for Disease Control and Prevention’s Emerging Infections Program (EIP).⁴ Since then, however, the US incidence of CDI has declined from an estimated 154.9 cases/100,000 population in 2011 to 143.6 cases/100,000 population in 2017. After accounting for changes in diagnostic testing, the adjusted incidence of CDI in 2017 was estimated to be 113.5 cases/100,000 population, which corresponded to an adjusted decrease of 24% in the incidence of CDI since 2011.⁴ The decrease in CDI incidence has been attributed to a steady decline in healthcare-associated CDI. The onset of the coronavirus disease 2019 (COVID-19) pandemic in 2020 appeared to have had minimal impact on the decreasing incidence of healthcare-associated CDI in the United States, with most multicenter studies reporting no change or a decrease in healthcare-associated or hospital-onset CDI rates compared to the prepandemic period.^5,6 In addition, between 2015 and 2020, there was a 56.8% decrease in the number of hospitalized CDI events, including hospital-onset CDI, that was reported to the US National Healthcare Safety Network (Fig. 2); this decrease corresponded to a 47.5% decline in the national CDI standardized infection ratio, which is a risk-adjusted measure of hospital-onset CDI.⁷ Since 2020, the annual number of reported hospitalized CDI events has remained stable.⁷ Several factors may have contributed to the decline in healthcare-associated CDI over the last decade, including improved infection control and prevention practices, antibiotic stewardship, a decrease in the prevalence of RT027/ST1, and changes in CDI diagnostic testing practices.⁴

In contrast, the estimates of the US burden of CA CDI, adjusted for factors such as testing practices, have not decreased over the last decade.⁴ In fact, since the initial reports of severe CDI cases in the community in the mid-2000s,⁸ the crude incidence of CA CDI has steadily increased over time, with the exception of 2020, when the crude incidence rate dropped, likely due to a decrease in outpatient health care utilization during the onset of the COVID-19 pandemic.⁹ Since then, the crude incidence rate of CA CDI has risen back to the prepandemic level. As of 2022, over half of the CDI cases reported through CDC’s EIP surveillance are CA, with the majority having had recent outpatient health care exposures. Meanwhile, hospital-onset and postdischarge CDI cases each account for only 19% to 20% of all reported cases (Fig. 3).

Similarly, the epidemiology of CDI in Canada has evolved since the initial outbreaks of RT027/ST1 in the early 2000s. From 2009 to 2015, the national rate of healthcare-associated CDI in Canada decreased from 5.9 to 4.3 cases/10,000 patient-days and was associated with a reduction in the prevalence of RT027/ST1.¹⁰ Subsequent surveillance during 2015 to 2019 showed an additional decrease of 23.8% in healthcare-associated CDI from 4.7 to 3.6 cases/10,000 patient-days.¹¹ Early on, between 2011 and 2014, there appeared to be an increasing trend in the national rate of hospitalized CA CDI.¹² However, between 2015 and 2019, the national rate of hospitalized CA CDI decreased by 18.8%.¹¹

Data from Mexico are included in a later section describing Latin America.

Since its initial spread from North America, RT027/ST1 has become less frequently detected in Europe in recent years, coinciding with a decline in CDI rates in some European countries. The concomitant decrease in CDI rates has been most striking in the United Kingdom, where rates in England declined by 77.0% from 107.6 cases/100,000 population in 2007–08 to 24.8 cases/100,000 population in 2013–14.¹³ Rigorous epidemiologic assessment of declining fluoroquinolone prescribing in England during this period, and its association with declines in the fluoroquinolone-resistant RT027/ST1, point to the important role of restrictions on fluoroquinolone prescribing over and above infection control in these declines.¹⁴ However, starting in 2021–22, rates have increased annually by 9.2%–11.3%, including increases in both healthcare-associated and CA CDI rates.¹³

Since 2016, 26 countries have reported hospital-based surveillance data to the European Center for Disease Prevention and Control. Data from 17 countries were available for 2018 to 2020, revealing the mean hospital incidence density of healthcare-associated CDI to be 2.8, 2.0, and 2.6 cases/10,000 patient-days in 2018, 2019, and 2020, respectively.¹⁵ The reported mean hospital incidence of CA CDI (which also included cases with unknown epidemiologic association) was 0.7 cases/10,000 patient discharges in 2018 and 2019 and 1.4 cases/1000 patient discharges in 2020, although the majority of the data in 2020 were contributed by Germany, which had a comparatively higher incidence of CA CDI than the other countries.¹⁵

In Australia, patient administrative data are used to monitor the national incidence of CDI. The overall rates of CDI diagnoses among hospitalized patients remained relatively stable from 2012 to 2019, with the average annual rate varying between 3.8 and 4.3 cases per 10,000 patient-bed days.¹⁶ However, during 2020 to 2021, large variations in quarterly rates of CDI were observed. Starting in the second quarter of 2021, CDI rates steadily increased, resulting in an average annual rate of 4.8 cases/10,000 patient-bed days, which was the highest reported since 2016.¹⁶ During 2020 to 2021, community-onset CDI accounted for over 80% of the hospitalizations with a CDI diagnosis.¹⁶ While reasons for these recent increases and prominence of the community in the epidemiology of C difficile in Australia are not clear, a close examination of infection control and antibiotic prescribing in ambulatory settings appear warranted. However, an in-depth look at the phylogenetics and resistome of major strains of circulating C difficile suggest the need to consider the role of emerging environmental and One Health sources of C difficile in Australia.¹⁷

Compared to the West, CDI incidence is less well characterized in Asia, where there is currently no national surveillance system for CDI. The few studies that have provided national or regional CDI data thus far are from East Asia. These include a multicenter study in Korea, where active surveillance for CDI was conducted over a 3-month period in 18 hospitals across the country during 2020 to 2021, resulting in an overall incidence of 7.04 cases/10,000 patient days.¹⁸ Using national health insurance claims data, the same investigators found the incidence of CDI in Korea had increased annually between 2008 and 2020, from 0.3 to 1.8 cases/10,000 patient days and from 6.9 to 56.9 cases/100,000 population.¹⁸ Similarly, in Hong Kong, a territory-wide study found that the crude incidence of CDI had increased substantially from 16.2 cases/100,000 population in 2006 to 49.4 cases/100,000 population in 2014.¹⁹ A subsequent study indicated the crude incidence of CDI in Hong Kong peaked at 56.9 cases/100,000 in 2017, followed by slight decreases during 2018 to 2019.²⁰ In a meta-analysis of 33 studies from 14 provinces in China published during 2017 to 2022, the pooled prevalence of CDI among symptomatic patients was 11.4%, with a higher prevalence in northern (13.6%) than southern China (11.0%).²¹ The pooled prevalence rates of healthcare-associated and CA CDI were 13.3% and 9.5%, respectively. In Japan, a study of patients aged ≥50 years in 9 hospitals in Tokyo during 2018 to 2020 revealed an incidence rate of 2.0 cases/10,000 patient-days and 97 cases/100,000 population per year.²² Extrapolation of this data to the national level indicated that greater than 57,000 hospitalized CDI cases occur annually in Japan in persons aged ≥50 years.²²

Little is known about the epidemiology of CDI in other parts of the world, where there may be less awareness of the organism and no formal surveillance in many low-income and middle-income countries. In Latin America, most of the studies on CDI have come from Brazil, where the prevalence of CDI among hospitalized patients varied between 0% and 41.7% during 2017 to 2019.^23,24 In a multicenter study in Mexico that included 16 hospitals from different regions, the incidence of CDI during 2016 to 2022 was 4.8 cases/10,000 patient days.²⁵ Although there are limited data on CA CDI in Latin America, one single-center study in Argentina reported that 34.4% among all CDI cases identified during 2014 to 2019 were CA.²⁶

Among studies conducted in Africa during 2015 to 2018, the prevalence of toxigenic C difficile in pediatric or adult patients with diarrhea generally ranged from 3.0% to 33.8%.²⁷ One outlier study in Kenya reported a high CDI prevalence of 92.4% among patients presenting with diarrhea over a 3-month period,²⁸ raising the possibility that the data might represent an outbreak.²⁷ A study in South Africa that used data from the national laboratory database reported an estimated crude national incidence of 53.9 cases/100,000 admissions in 2016 to 2017.²⁹ More recently, a multicenter study of hospitals from 3 different provinces in South Africa reported median CDI rates ranging from 0.3 to 5.8 cases/10,000 patient-days during 2017 to 2020, with an increase in rates in 2020 compared to 2019.³⁰

Extensive studies of the innate immunity to CDI have uncovered pivotal roles played by various immune cells and cytokines.³¹ In addition, early studies of antibody responses to CDI, as well as later studies evaluating the use of monoclonal antibodies, suggest that an adequate humoral response to C difficile toxins A and B can protect against initial and recurrent disease.³² Thus, impairment of the immune system, as seen in older age and in immunocompromised individuals, whether due to a primary or secondary immunodeficiency, can increase the risk of CDI. Among immunocompromised patients, prevalence or rates of CDI have ranged from 6% to 33% in the hematology-oncology population, 0.6% to 23% in solid organ transplant recipients, and 7.1 to 8.3 cases/1000 patient-years in patients with human immunodeficiency virus.³³

Disruption of the homeostasis of the gastrointestinal microbiome can lead to changes in the composition of commensal organisms as well as the metabolome (especially bile acids), pH, and epithelial mucus, all of which can allow C difficile, if present, to proliferate and cause symptomatic infection.³⁴ Antibiotic use is a well-recognized CDI risk factor due to its profound and long-lasting impact on the microbiome. Almost every antibiotic has been implicated, but those with broad-spectrum activity or those to which some C difficile strains are highly resistant appear to have the highest risk for CDI.¹ These include clindamycin, cephalosporins, fluoroquinolones, and penicillins, whereas risk of CDI is lower with macrolides and sulfonamides and lowest with tetracyclines.^35,36 Risk of CDI also varies within classes of antibiotics. For example, amoxicillin-clavulanate has the highest level of risk within its class, supporting the notion that penicillin-beta-lactamase combinations should be regarded differently from other penicillins.^35,36 Risk of CDI also differs by duration of antibiotic exposure, with a 14-day course on average having 27% more risk compared to a 7-day course, while a 5-day course has 9% less risk.³⁶ However, risk is greater with longer durations of high-risk antibiotics than low-risk antibiotics. Moreover, exposure to multiple antibiotics and cumulative exposures over 2 to 3 months increase the risk of CDI more than a short course of a single antibiotic.³⁷ While direct antibiotic exposure is associated with an increased risk of CDI, studies have found that even antibiotic use among other patients in the same ward or a prior occupant of the same hospital room can independently increase an individual’s risk for CDI.^38,39 Although the specific mechanisms underlying the herd effects of antibiotics have not been thoroughly studied, it is likely that antibiotics may promote C difficile shedding, resulting in a higher environmental burden of the organism and an increased risk of acquisition in surrounding and subsequent patients.

While antibiotics are well-studied, there is increasing recognition that other medications can also disrupt the intestinal microbiome. Still, thus far, only a few of these have been linked to an increased risk of CDI. These include acid-suppressing medications, with more evidence supporting an association with proton pump inhibitors (PPIs) than histamine-2 receptor blockers.⁴⁰ Although conflicting data exist, a review of meta-analyses and systematic reviews conducted between 2000 and 2020 found a significant association between PPIs and CDI, with the majority of studies reporting 50% to 100% higher odds of CDI among PPI-users compared to nonusers.⁴⁰ In addition, the hypochlorhydria effect of PPIs has been demonstrated to promote C difficile sporulation,⁴¹ as well as alter the taxa involved in colonization resistance to C difficile.⁴² On a population level, it is possible that the effect of PPIs on CDI risk might be greater than that of antibiotics since PPIs are more commonly used and often given on a chronic basis for maintenance therapy compared to antibiotics that are usually given for a shorter duration.⁴³ Less data are available regarding other medications, with a few studies suggesting an increased risk of CDI from antidepressant use,⁴⁴ although the mechanisms for this are not clear and could include not only alteration of the microbiota but also dysregulation of the immune system.⁴⁵

Certain patient characteristics or underlying conditions can also disrupt the microbiome. For example, while older age can increase the risk of CDI through waning immunity, there are data indicating that older age is also associated with changes in the intestinal microbiome, including a reduced diversity of species that can result in less inhibitory effect on C difficile growth.⁴⁶ Other conditions include inflammatory bowel disease, where affected patients have a higher incidence of developing CDI compared with the general population.⁴⁷ In addition to the increased risk from frequent immunosuppressant use, patients with inflammatory bowel disease have an altered microbiome due to the underlying pathogenesis of the disease, which can potentially facilitate C difficile growth. This is supported by the finding that patients with ulcerative colitis have a significantly higher risk of CDI than those with Crohn’s disease, who generally have less extensive colonic disease.⁴⁷

The risk of a patient acquiring C difficile depends on the duration of exposure and the burden of the organism in the surrounding area, which includes not only the extent of the environmental contamination with C difficile, but also the number of patients in each unit who are actively infected or colonized with C difficile. In addition, risk of C difficile transmission is influenced by individual behavior or practices. These consist of lapses in infection control and prevention practices by facility staff that could facilitate C difficile transmission between patients, including suboptimal hand hygiene and/or glove or gown use, as well as nonadherence to recommended cleaning and disinfection procedures for patient rooms and shared medical equipment or devices. Additional practices that could increase transmission risk include the sharing of toilets or bathrooms or the use of lidless toilets.⁴⁸ While these practices can be independently associated with C difficile transmission, their level of risk may vary according to the environmental burden of the organism and the level of colonization pressure (ie, proportion of patients in the unit colonized with C difficile).

Colonization pressure is a known risk factor for several healthcare-associated infections⁴⁹; however, it has been less studied in the context of CDI, as screening for asymptomatic C difficile carriers is not routinely performed. However, 1 early study of intensive-care unit patients indicated that C difficile colonization pressure was an independent risk factor for CDI,⁵⁰ and in recent years, several studies have demonstrated that asymptomatic carriers can serve as a reservoir for CDI. For instance, 1 study found that incident CDI cases in a hospital were as frequently linked to transmission from asymptomatic carriers (29%) as symptomatic patients (30%).⁵¹ Another study found that among healthcare-associated CDI cases identified in a hospital and its affiliated long-term care facility, 19% were linked to long-term care facility residents with CDI or asymptomatic carriage of toxigenic C difficile.⁵² All of the asymptomatic carriers linked to transmission in the study had a relatively high burden of C difficile based on the colony count from their perirectal swab and evidence of shedding from positive groin, skin, and/or environmental cultures.

While C difficile contamination of the hospital environment has been well-described and linked to hospital-onset CDI,⁵³ limited data indicate that outpatient settings can also be contaminated with C difficile and potentially serve as a reservoir for CA CDI. In 1 study, point-prevalence cultures of high-touch surfaces revealed that 14% of outpatient clinic and emergency department rooms were contaminated with C difficile.⁵⁴ A more recent study found a lower rate of C difficile environmental contamination in outpatient clinics, ranging from 0% to 5% of high-touch surfaces.⁵⁵ Prevalence of C difficile contamination likely varies by outpatient setting type and is influenced by patient volume, frequency of patient turnovers, and the ability of facility staff to perform adequate environmental cleaning and disinfection between patient encounters. Studies have reported that as high as 82% of CA CDI cases have had an outpatient health care encounter in the preceding 12 weeks.⁵⁶ Thus, it is conceivable that some cases could have acquired C difficile from their recent outpatient encounter, particularly if there was prolonged exposure and contact with providers and the health care environment. In a large, multisite study, an emergency department visit in the preceding 12 weeks, independent of receiving antibiotics, was significantly associated with CA CDI, suggesting that emergency departments could be an environmental source of CDI.⁵⁶

Increasing evidence suggests that a large portion of hospital-onset CDI are caused by strains that originated from the community, either acquired before hospital admission or from another patient who is asymptomatically colonized.⁵⁷ In addition, the increasing incidence of CA CDI in some countries has prompted greater attention to community sources of CDI. Studies have shown that infants and infected household members, as well as farm animals, could be a source of CA CDI.^58–60 The use of whole-genome sequencing has revealed potential transmission of C difficile strains RT078 and RT046 between humans and pigs in Europe^61,62 and RT014 and RT001 between humans and livestock in Nigeria.⁶³ Toxigenic C difficile has also been isolated from soil,⁶⁴ water,⁶⁵ and various types of food,^66,67 including retail meats and root vegetables. Although no direct foodborne link has been established for CDI, a European study identified 2 distinct patterns of C difficile spread: one reflecting healthcare-associated transmission, and the other suggesting transmission via other routes or sources, such as the food chain.⁶⁸