(as is often true during outbreak investigations) or the incidence of certain clinical signs may be monitored. Despite the difficulty, if no attempt is made to standardize the grounds for a diagnosis or create a working definition, it is hard (if not impossible) to interpret changes in disease frequency. The key is to standardize as much as possible what staff diagnose as a particular disease or identify as a clinical sign. The definitions need not be perfect, only consistently applied.
The collection of the date of the first diagnosis is essential to mark the onset of each new case so that incidence measures can be calculated for specific time periods. For diseases that can occur more than once in an individual animal (e.g. upper respiratory tract infections), only the first episode during a particular timeframe is counted as a new or incident case. If the medical team has an interest in reoccurring illnesses, the rate of second occurrences can be calculated and reported separately. If second (or other) occurrences are monitored during a time period of interest, the denominator includes only animals with a first occurrence in that timeframe. Fortunately, since most animals do not reside in most shelters long enough to experience second infections of most diseases, the rate of second infections is usually ignored. An exception may be in sanctuaries.
3.2.4 Data Collection, Analysis, Interpretation, and Communication
Disease surveillance takes time and resources to do well. An effective surveillance program must be valued, planned, and well‐executed. Written protocols governing what, where, when, and by whom each component will be performed are essential.
Several staff members are usually involved in data collection during an animal's passage through the shelter system. For infectious disease surveillance, shelter intake (for denominators) and medical data (for numerators) are obviously needed, but data related to movement, outcomes, daily observations and other events could also be important to address questions that arise from surveillance. Quality and completeness of all relevant data are key components. Everyone involved with data collection must be trained and held accountable for providing good data. Without explicit protocols, staff may be unsure of how, what, when, and where to collect specific pieces of information.
How and by whom the data will be routinely analyzed should be clear. This includes the metrics (e.g. incidence, mortality) and subgroups of animals to monitor, the trends to track, and any other metrics that are important to the shelter's medical‐related goals.
Since disease risk frequently varies by age group, source (e.g. stray), and over time, separate incidence rates should be calculated for each of these factors. Data can contradict common beliefs. For example, after reviewing Figure 3.1, medical staff members were surprised that the incidence (or risk) of feline upper respiratory tract disease (URTD) was actually higher in the fall and winter months than in the spring and summer. Further analyses demonstrated that this was true among both kittens and adult cats. The staff did see more sick cats in the spring and summer, but the cat intake was also much higher. With these results, the shelter heightened its adherence to its preventive protocols in the fall and winter. Similarly, surveillance data can confirm suspicions and provide evidence to change behavior and shelter protocols. A veterinarian had the clinical impression that the survival rate among fostered un‐weaned kittens could be improved by making changes to their management in foster homes; foster care providers were resistant to change. The data (see Figure 3.2) demonstrated a survival rate of only 71%. The providers were unaware of this rate, and they also found it unacceptable. The veterinarian updated protocols and made presentations to explain the basis for the changes. Over the next several years, survival rates increased to over 90%.
Figure 3.1 Incidence rate of URTD among cats by season.
Figure 3.2 Annual survival rates of un‐weaned kittens in foster care 2011–2014.
Routine monitoring and analysis of data augment the daily observations of the medical staff, adding to their understanding of the state of the health of the population. When routinely incorporated into population care, data surveillance can improve the health and welfare of shelter animals.
3.2.5 Frequency of Review, Interpretation and Communication
A key component of disease surveillance is regular data review and communication (Anderson 1982; Nelson and Williams 2007). Regular review by medical staff facilitates prompt recognition of changes in disease frequency and heightens responsiveness to these changes. This facilitates the timely achievement of disease‐related goals and enables the medical staff to inform the administration and other shelter staff of health changes in a timely manner. No less than monthly reviews of the frequency of some diseases (e.g. upper respiratory tract infections) at regular staff meetings are warranted. Monthly reviews have the added benefit of maintaining interest in disease surveillance and attention to the collection of quality data. It is powerfully motivating for staff to realize that their adherence to protocols was responsible for falling disease rates and the achievement of other goals.
The timing of some analyses can be pre‐specified, but other analyses may need to occur as the data suggests new questions. Frequency of monitoring is usually related to the incidence and severity of the disease, and whether the shelter has specific goals related to that disease. Generally speaking, high‐incidence diseases (e.g. feline URTD) will be monitored more frequently (e.g. monthly) than those of lower frequency. Disease outbreaks also trigger more frequent monitoring.
Interpretation of the results should include staff that may have insight into a particular analysis. The medical staff should obviously be involved but, in many instances, the shelter manager, kennel personnel or volunteers may also have important insights into why disease is manifesting as it is. For example, the interpretation of data relating to a rise in the mortality of fostered kittens might involve the shelter manager, veterinarian, executive leadership, foster‐care providers and others empowered to make protocol changes based on that data.
The ability to interpret and communicate findings based on the data is heavily influenced by the data presentation; the presentation can obscure or enhance the message it contains. It is beyond the scope of this chapter to make recommendations for effective data presentation, but references are provided (Knaflic 2015; Tufte 2001). A consistent recommendation in these references is to keep presentations simple and focused on the message that needs to be conveyed; it is important to avoid ostentatious presentations that obscure that message.
3.2.6 Data Quality and Administrative Buy‐In
Since disease surveillance requires the commitment of staff throughout a shelter, members of the shelter administration and management staff must believe in and support surveillance efforts. Staff training and incentives enhance the collection of quality data. When staff members understand the rationale for, and the importance of, complete and accurate collection of information, data quality rises. Resources are wasted in the collection of poor‐quality data, and analyses of that data can be misleading. Summary reports should be shared with staff on a regular basis. A review of the analyses facilitates discussions aimed at reducing disease and provides an opportunity to celebrate staff contributions to declining disease rates!
3.2.7 Recording Changes Affecting the Surveillance Program
Changes that can have an impact on disease frequency occur routinely in shelters. For example, turnover