To better understand the role of plasmids and their importance in the endemic antibiotic resistance of Enterobacteriaceae, we began a prospective study of our combined medical intensive care/coronary care unit. An initial culture survey of the patients, ward staff, and environment was followed by a prospective sampling of 139 consecutive new admissions at the time of admission to the unit, and at regular intervals thereafter for the remainder of their stay. All cultures were planted on agar-containing gentamicin. Of the 147 patients studied, 12 (8.2%) were colonized with 20 strains of gentamicin-resistant gram-negative bacilli (GRGNB) at 29 sites. An additional four GRGNB were isolated from the environment. Of the 24 GRGNB strains, 7 (29%) Enterobacteriaceae carried plasmids shown to carry the gentamicin-resistance determinant. Plasmids were further characterized by restriction endonuclease digestion profiles of plasmid DNA purified from E. coli C600 transconjugants or transformants. A 93 kb plasmid introduced to the unit by a Serratia liquefaciens colonizing a patient transferred from another area in the hospital was identical to 93 kb plasmids carried by a C. amalonaticus and an E. aerogenes subsequently colonizing another patient on the unit. A 60 kb plasmid, first isolated from a S. marcescens colonizing a sink drain was later isolated from an E. cloacae colonizing a patient. Our results indicate that spread of specific R-plasmids may be one mechanism for dissemination of antibiotic resistance on our MICU/CCU.

We identified 29 episodes of suppurative thrombophlebitis in 27 patients admitted to a large general hospital between May 1980 and May 1984. In 25 patients, the intravenous cannulae had been in place for more than 3 days. Streptococcus faecalis, Pseudomonas aeruginosa or one of the Enterobacteriaceae were implicated in 14 patients. All these patients had recently undergone abdominal surgery or had a major intraabdominal inflammatory process at the time they developed thrombophlebitis. The remaining 13 patients were infected with Staphylococcus aureus, other grampositive cocci or Candida species. Only two of these had an active abdominal process at the time of their infection (x2 = 16.08, P<0.001). There is an apparent association between phlebitis caused by enteric organisms and active intra-abdominal pathology. There were two deaths related to delayed or deferred surgery. Suppurative thrombophlebitis is a lethal, preventable nosocomial infection that requires urgent surgical intervention.

Infection of the urinary tract is acknowledged to be the most common hospital infection, associated closely with the presence of an indwelling catheter. Patients are catheterized for a variety of reasons and for different periods of time, ranging from postoperative catheterization of a few days following urological surgery to the long-term catheterization over many months or years of patients who are unfit for operation, with spinal injuries or neuropathic bladders. The extent of the problem includes patients returning home infected, requiring catheterization before readmission to the hospital or needing nursing at home with a long-term catheter. The risks of infection and its complications as well as methods of control may differ between each group.

It is of fundamental importance to distinguish between patients who came to surgery with an existing infection and those with sterile preoperative urine. Other factors include recent previous catheterization and, importantly, the length of time the catheter is inserted. For the patient with sterile preoperative urine, postoperative bacteriuria seldom causes severe symptoms and can be treated with antibiotics or left to clear spontaneously after removal of the catheter. A minority of patients suffer consequences of their infection especially when bacteriuria starts before the catheter is removed postoperatively. Catheter removal often causes transient bacteremia, also induced by instrumentation or operation on infected urine, which may lead to serious complications, particularly of septicemia. In our experience in Bristol, about 1 in 4 patients admitted for urological surgery already has infected urine as defined by the presence of ≥105 bacteria/ml midstream urine or ≥=104 bacteria/ml catheter urine. Of those admitted with infected urine, 3 in 4 have a catheter already inserted compared with only 1 catheterized patient in 4 admitted to operation with sterile urine.

In the fall of 1983 an outbreak of hand-foot-and-mouth disease occurred in a large urban hospital. The outbreak began among surgical nurses confined to one area of the operating suite and appeared to spread by direct contact to personnel working in the other areas. No cases were diagnosed among hospital patients. There was no evidence to support an ongoing community epidemic as only three of 98 physicians surveyed had seen hand-foot-and-mouth disease in their practices, and none occurred among family members of 94 unaffected employees. The outbreak resulted in 82 lost workdays at an estimated cost to the hospital of $5,676. Existing infection control guidelines do not address the issue of transmission of hand-foot-and-mouth disease from hospital personnel to patients; our experience suggests that the risk of transmission, at least in the operating suite environment, may be greater for other personnel than for patients.

Much attention has been given to ventilators, respiratory tubing and humidifiers as sources of bacteria causing pneumonias in intensive therapy unit (ITU) patients. However, efficient decontamination and safe storage of ventilators and humidifiers between patients, with frequent changing of respiratory tubing and the use of Chlorhexidine in humidifier reservoirs, are now undertaken in most ITUs. In addition, chest aspiration, which is an important route for bacterial entry to the chest, is performed by trained ITU staff, who are well aware of this danger and wear gloves and use meticulous techniques. As a result, it can be argued that infection of the lungs in ITU patients from these sources and routes has reached an “irreducible minimum.”

Unfortunately, many patients in ITUs continue to develop pneumonias. In my own experience patients yielded Pseudomonas in tracheal aspirates as frequently after the introduction of these techniques, together with the insertion of efficient bacterial filters, between the patient and the ventilator in both the inspiratory and expiratory lines, as before their adoption. What other sources and routes of infection of the lungs should be considered?

The adult health care provider who is susceptible to varicella zoster virus (VZV) represents a risk to her or himself and to patients. Nineteen percent of employees at this Children's Hospital had no or uncertain prior experience with VZV, and of these, 28% were found to be VZV susceptible, representing 5% of the total population of 2,730 hospital employees. During the 12 months of study, six of the potential 135 to 137 VZV-susceptible individuals acquired varicella. VZV-susceptible health care providers should be aware of their potential to both acquire and transmit VZV.

Studies of the risk factors for bacteriuria associated with the use of temporary indwelling urinary catheters have focused on gram-negative bacilli and enterococci, the predominant causes of nosocomial urinary tract infection. During a prospective study of 747 catheterized patients, bacteriuria with ≥1,000 bacteria/mL were found in 96 patients. Coagulase-negative staphylococci (CNS) accounted for 19% of the episodes of bacteriuria, and S. epidermidis was the most frequent species of CNS. Significant risk factors for bacteriuria were a positive meatal culture for CNS, non-surgical illness, and no antibiotic therapy during catheterization. In contrast to the results of this and earlier studies that females were at greater risk factor for catheter-associated bacteriuria with gram-negative bacilli and enterococci, this was not a risk factor for acquisition of CNS. Possibly this is due to the more similar prevalence of meatal colonization with CNS of males and females compared to the striking difference in prevalence of meatal colonization with gram-negative bacilli and/or enterococci of males and females.

Admission of a patient with group A streptococcal cellulitis and bacteremia to the intensive care unit of a tertiary care teaching hospital was followed by two subsequent cases of group A streptococcal bacteremia with pneumonia in the unit. All streptococcal isolates were the same M- and T-type. Endotracheal intubation with respiratory ventilation was a risk factor for disease acquisition. The characteristics of onset of the two nosocomially acquired cases suggested that a staff member may have been, at least transiently, a streptococcal carrier, but no such carrier was identified. No further cases occurred subsequent to a period when all patients in the unit received antibiotics effective for group A streptococcal therapy.