To examine a comprehensive approach for preventing percutaneous injuries associated with phlebotomy procedures.

From 1993 through 1995, personnel at 10 university-affiliated hospitals enhanced surveillance and assessed underreporting of percutaneous injuries; selected, implemented, and evaluated the efficacy of phlebotomy devices with safety features (ie, engineered sharps injury prevention devices [ESIPDs]); and assessed healthcare worker satisfaction with ESIPDs. Investigators also evaluated the preventability of a subset of percutaneous injuries and conducted an audit of sharps disposal containers to quantify activation rates for devices with safety features.

The three selected phlebotomy devices with safety features reduced percutaneous injury rates compared with conventional devices. Activation rates varied according to ease of use, healthcare worker preference for ESIPDs, perceived “patient adverse events,” and device-specific training.

Device-specific features and healthcare worker training and involvement in the selection of ESIPDs affect the activation rates for ESIPDs and therefore their efficacy. The implementation of ESIPDs is a useful measure in a comprehensive program to reduce percutaneous injuries associated with phlebotomy procedures.

To estimate the frequency of, and assess risk factors for, percutaneous, mucous membrane, and cutaneous blood contacts sustained by healthcare workers (HCWs) during the delivery of infusion therapy and the performance of procedures involving sharp instruments in the home setting.

Prospective surveillance of percutaneous, mucous membrane, and cutaneous blood contacts.

Eleven home healthcare agencies in the United States and Canada from August 1996 through June 1997.

HCWs who provided home infusion therapy or performed procedures using hollow-bore needles and other sharp instruments in the home setting.

Each participating worker recorded information about the procedures performed and blood contacts experienced during each of his or her home visits for a 2- to 4-week period using standard questionnaires. HCWs also completed questionnaires regarding job duties, reporting of previous occupational blood contacts, and their use of protective barriers in the home setting.

Participating HCWs provided information about 33,606 home visits. A total of 19,164 procedures were performed during 14,744 procedure visits. Fifty-three blood contacts occurred during these visits, for a blood-contact rate of 2.8 blood contacts per 1,000 procedures and 0.6 percutaneous injuries per 1,000 procedures with needles or lancets. Gloves were worn for 52%, masks for 5%, gowns for 3%, and protective glasses or goggles for 2% of all procedure visits. HCWs used barriers for 53% of visits during which at least 1 procedure was performed and for 27% of other visits.

HCWs involved in home health care are at risk for blood contact. Infection control barrier use was low in our study. The majority of skin contacts could have been prevented by glove use.

To study the epidemiology and preventability of blood contact with skin and mucous membranes during surgical procedures.

Observers present at 1,382 surgical procedures recorded information about the procedure, the personnel present, and the contacts that occurred.

Four US teaching hospitals during 1990.

Operating room personnel in five surgical specialties.

Numbers and circumstances of contact between the patient's blood (or other infective fluids) and surgical personnel's mucous membranes (mucous membrane contacts) or skin (skin contacts, excluding percutaneous injuries).

A total of 1,069 skin (including 620 hand, 258 body, and 172 face) and 32 mucous membrane (all affecting eyes) contacts were observed. Surgeons sustained most contacts (19% had ≥1 skin contact and 0.5% had ≥1 mucous membrane-eye contact). Hand contacts were 72% lower among surgeons who double gloved, and face contacts were prevented reliably by face shields. Mucous membrane-eye contacts were significantly less frequent in surgeons wearing eyeglasses and were absent in surgeons wearing goggles or face shields. Among surgeons, risk factors for skin contact depended on the area of contact: hand contacts were associated most closely with procedure duration (adjusted odds ratio [OR], 9.4; ≥4 versus <1 hour); body contacts (arms, legs, and torso) with estimated blood losses (adjusted OR, 8.4; ≥1,000 versus <100 mL); and face contacts, with orthopedic service (adjusted OR, 7.5 compared with general surgery).

Skin and mucous membrane contacts are preventable by appropriate barrier precautions, yet occur commonly during surgery. Surgeons who perform procedures similar to those included in this study should strongly consider double gloving, changing gloves routinely during surgery, or both.