Describe hospital infection surveillance and the investigation of an outbreak. Add a note on the infection control committee.

Surveillance

• Rates must be expressed against a denominator that reflects exposure, or comparison is meaningless:
• Device-associated rate = infections ÷ device-days × 1,000. Using admissions or bed-days instead would make a unit look better simply for having fewer catheters — which is the right thing to do, but it must be captured as reduced device use rather than hidden
• Device utilisation ratio = device-days ÷ patient-days, which measures how much a unit uses devices and is itself a target for improvement
• Surgical site infection rate is stratified by wound class and by the NNIS or ASA risk index, because a contaminated emergency laparotomy and a clean elective hernia repair are not comparable
• Standard case definitions are essential — without them, a falling rate may simply mean stricter definitions or less diligent case-finding

Outbreak Investigation

• 1. Establish the existence of an outbreak — compare with the baseline rate, and exclude pseudo-outbreaks caused by a new laboratory method, better case-finding, a contaminated reagent or a change in definition → 2.
• Confirm the diagnosis — verify the laboratory results and save the isolates for later typing → 3.
• Define a case and find cases — an explicit case definition by person, place and time, applied consistently
• then active case-finding → 4.
• Descriptive epidemiology — describe by time (an epidemic curve, whose shape distinguishes a point source from person-to-person spread), place (a spot map by ward, bed and theatre) and person (age, procedure, device, ward, and which staff were involved) → 5.
• Form a hypothesis about the source and the mode of transmission → 6.
• Test it analytically — a case-control study is usual in a hospital outbreak, since cases are few and exposures many
• a cohort study where the exposed group is defined → 7.
• Environmental and microbiological investigation — targeted sampling guided by the hypothesis, and typing of isolates by pulsed-field gel electrophoresis, MLST or whole-genome sequencing to establish whether the strains are truly the same → 8.
• Implement control measures — and do so as soon as a plausible source is suspected, without waiting for the analysis to be complete → 9.
• Evaluate by continued surveillance, and 10.
• Communicate — a written report, feedback to staff, and notification where required
• Control measures come before certainty, and this is the point at which investigations most often fail. The purpose of the investigation is to stop the outbreak, not to publish it, and a reasonable measure taken early is worth more than a perfect answer arrived at late
• Do not undertake routine environmental sampling — swabbing walls, floors and air without a hypothesis wastes resources, generates uninterpretable results and finds organisms that are always present. Sampling is done to test a specific hypothesis
• Beware the pseudo-outbreak — a cluster of positive cultures with no clinical disease, caused by contaminated collection equipment, a contaminated reagent or laboratory cross-contamination; the clue is that the patients are not ill

The Infection Control Programme

• The hospital infection control committee (HICC) — a multidisciplinary committee chaired at senior level, including the microbiologist, clinicians, surgeon, nursing superintendent, pharmacist, hospital administrator and engineer. It sets policy, reviews surveillance data and authorises action
• The infection control team (ICT) — the working arm, comprising an infection control officer (usually the microbiologist) and infection control nurses, who do the surveillance, education and daily liaison. A ratio of about one infection control nurse per 250 beds is the usual recommendation, and staffing this post is the single most important organisational decision
• Functions — surveillance with feedback; writing and updating policies (hand hygiene, isolation, device bundles, antibiotic policy, waste, sterilisation); education and training of all grades including cleaning and waste staff; outbreak investigation; audit; staff health and immunisation; and advising on new construction and renovation
• Antimicrobial stewardship is usually run alongside, sharing the same data and often the same people
• Accreditation — NABH in India and JCI internationally require a functioning infection control programme, which has been a significant driver of improvement in Indian hospitals

Common Hospital Outbreaks and Their Sources

Applied Aspects

• Report a cluster early rather than waiting to be sure; the cost of investigating a false alarm is small and the cost of a delayed outbreak is not
• Save the isolates when an unusual organism appears; typing is impossible retrospectively if the plates have been discarded, and without typing an outbreak cannot be confirmed or excluded
• Feed the data back to the clinical teams in a form they can act on — unit-level rates, trends, and comparison with previous periods; data sent only to administration changes nothing
• Do not use surveillance data punitively; the moment reporting is perceived as blame, case-finding deteriorates and the rates fall for the wrong reason
• Post-discharge follow-up is necessary for surgical site infection, or the reported rate will be a fraction of the true one and will mislead everybody
• Involve the engineers in outbreaks involving water, air conditioning or construction; Legionella, Pseudomonas and Aspergillus outbreaks are solved by plumbing and ventilation rather than by antibiotics
• Implement control measures before the investigation concludes; the purpose is to stop the outbreak, and a reasonable early measure beats a perfect late answer
• Recognise the pseudo-outbreak — positive cultures without ill patients point to contaminated reagents, collection equipment or laboratory cross-contamination rather than to a real cluster
• Assess construction and renovation in advance; an infection control risk assessment before building work begins prevents aspergillosis outbreaks that are very difficult to control once started
• Denominators matter — a unit that reduces catheter use will appear unchanged on a per-admission rate and improved on a per-device-day rate, and both numbers are needed to understand what happened
• Use standard case definitions, or an apparent improvement may simply reflect stricter criteria or less thorough case-finding
• An epidemic curve distinguishes a point source from spread — a sharp single peak suggests a common vehicle, successive waves suggest person-to-person transmission, and the shape guides the investigation before any culture is available
• Whole-genome sequencing has changed outbreak investigation, frequently showing that apparently linked cases are unrelated and that unsuspected cases belong together
• Fund the infection control nurse post; it is the single organisational decision most consistently associated with lower infection rates
• Targeted surveillance beats comprehensive surveillance where resources are limited, because data that nobody has time to analyse and feed back changes nothing
• Point prevalence surveys are a good starting point for a hospital with no surveillance at all — cheap, quick, and sufficient to establish that a problem exists and to justify a programme
• The SENIC finding still holds — surveillance with feedback and a trained infection control nurse reduces infection by about a third, and no single technical intervention matches that