Surveillance 4 (Data
interpretation )
,Generation, analysis and validation of surveillance data :
• Most generation of data is from available resources, such as admissions records,
questionnaires, interviews, medical records, public health reports, and laboratory
reports.
• Analysis should be done in a timely manner, especially important for the
detection of outbreaks. The type of analysis will depend upon the expected
outcomes and the purpose of surveillance.
• Validation of data is an ongoing process. All steps in the generation and analysis
of data should be reviewed regularly, especially when threshold rates have been
exceeded. If, for example, there is an apparent outbreak of antibiotic resistant
bacteria detected sputum cultures, then the procedures for obtaining the
specimens as well as laboratory procedures need to be validated to ensure that
the outbreak is not a pseudo-epidemic caused by faulty lab procedures or other
deviations in infection control.
,Preparing periodic reports of analyzed data
• Reports should be generated by the infection disease professional on a regular
basis, which may vary from monthly to quarterly or even annually, depending
upon the size of the institution and the population numbers or device days.
• Statistics must include adequate denominator data for meaningful analysis, and
this can require a longer period of time. Specific data about individual patients or
healthcare workers are often protected by laws regarding privacy, so information
about individuals cannot be disseminated unless anonymity can be assured.
• Reports to individual physicians about their own infections rates should be
provided confidentially and comparison rates done without identifying physicians.
• Reports are usually presented to the infection control committee, but reports
should also be presented to staff in areas of survey. Thus, if a study involved an
ICU, the ICU staff and physicians should be aware of the study results so that they
can evaluate the effectiveness of infection control procedures or institute
preventive methods.
, Comparison of rates
• Reports of hospital-acquired infections are often used as a basis of
comparison between one facility and others or one department in a
facility, such as an ICU, and another, such as a transplant unit. Even in-
house comparisons must be interpreted carefully because a higher rate of
infection does not always mean patients are at increased risk. Numerous
factors must be accounted for if a comparison is to be meaningful:
- Definitions must be uniform and consistent, following CDC definitions or
specific definitions that have been developed for the population at risk or
facility.
- Protocols for data collection should be uniform so that data is collected in
the same way in different units, and case finding should be consistent and
accurate. There should also be consistency in obtaining supporting laboratory
tests.
- Risk factors should be similar or results stratified to account for differences
in risk factors.
interpretation )
,Generation, analysis and validation of surveillance data :
• Most generation of data is from available resources, such as admissions records,
questionnaires, interviews, medical records, public health reports, and laboratory
reports.
• Analysis should be done in a timely manner, especially important for the
detection of outbreaks. The type of analysis will depend upon the expected
outcomes and the purpose of surveillance.
• Validation of data is an ongoing process. All steps in the generation and analysis
of data should be reviewed regularly, especially when threshold rates have been
exceeded. If, for example, there is an apparent outbreak of antibiotic resistant
bacteria detected sputum cultures, then the procedures for obtaining the
specimens as well as laboratory procedures need to be validated to ensure that
the outbreak is not a pseudo-epidemic caused by faulty lab procedures or other
deviations in infection control.
,Preparing periodic reports of analyzed data
• Reports should be generated by the infection disease professional on a regular
basis, which may vary from monthly to quarterly or even annually, depending
upon the size of the institution and the population numbers or device days.
• Statistics must include adequate denominator data for meaningful analysis, and
this can require a longer period of time. Specific data about individual patients or
healthcare workers are often protected by laws regarding privacy, so information
about individuals cannot be disseminated unless anonymity can be assured.
• Reports to individual physicians about their own infections rates should be
provided confidentially and comparison rates done without identifying physicians.
• Reports are usually presented to the infection control committee, but reports
should also be presented to staff in areas of survey. Thus, if a study involved an
ICU, the ICU staff and physicians should be aware of the study results so that they
can evaluate the effectiveness of infection control procedures or institute
preventive methods.
, Comparison of rates
• Reports of hospital-acquired infections are often used as a basis of
comparison between one facility and others or one department in a
facility, such as an ICU, and another, such as a transplant unit. Even in-
house comparisons must be interpreted carefully because a higher rate of
infection does not always mean patients are at increased risk. Numerous
factors must be accounted for if a comparison is to be meaningful:
- Definitions must be uniform and consistent, following CDC definitions or
specific definitions that have been developed for the population at risk or
facility.
- Protocols for data collection should be uniform so that data is collected in
the same way in different units, and case finding should be consistent and
accurate. There should also be consistency in obtaining supporting laboratory
tests.
- Risk factors should be similar or results stratified to account for differences
in risk factors.
