Consensus on which indicators are important for older adult fall prevention in Ontario and their order of priority was reached for both the MS and PH groups. The results of each group’s ranking varied significantly, highlighting the differing functions of the sectors represented in each group and therefore, the specific indicators that would benefit their work. Both groups ranked the rate of ED visits due to a fall high, likely reflecting the importance for both groups to understand and have the ability to present the population burden of fall-related injuries in their area. Typically, indicators are developed to provide monitoring and evaluation of a system. Future work including a continued process across other sectors could arrive at a parsimonious set of indicators that include MS and PH as relevant players in the system supporting older adult falls prevention, however at this stage, there are still discrepancies in the priority indicators for each group.
Public health indicated the proportion of falls by place of occurrence/activity as their highest ranked indicator, after ED visit and hospitalization indicators. This is due to its use in identifying high risk places and activities that can guide public health intervention planning. Currently, the most widely collected and accessible administrative data sources that could be used to populate this indicator are the National Ambulatory Care Reporting System (NACRS) and the Discharge Abstract Database (DAD), that contain information on all ED visits and hospitalizations in Ontario. While these data sources include ICD-10 codes that describe the external cause of both intentional and unintentional injury (V00-Y33), as well as a code describing the place of occurrence of the injury (U98), these alone do not provide sufficient detail to inform intervention planning. For example, the range of external cause codes for falls (W00-W19) broadly categorizes the cause of the fall (e.g. W00- fall on the same level involving ice and snow), but provides no additional information on how the fall occurred. Similarly, the codes that describe where the injury occurred are broad and limited (e.g. U98.1 residential institution, U98.9 unspecified place of occurrence) and do not include any specific information on where within that location the injury occurred. While this information would provide PHUs with more detail about the falls in their population than they currently have access to, there may be a need to access other data sources with more detailed place and activity information to maximize the utility of this indicator for users.
The PH group ranked the rate of hospitalizations due to a fall in second place; ranked 5th by the MS group. This may reflect the importance for PH to capture falls that result in more serious injuries and require hospitalization. This is additionally supported by the PH group’s high average ranking of the rate of serious fall-related injury hospitalizations indicator, as this indicator further distinguishes the most severe fall-related injuries, and is a more reliable reflection of serious injury than hospitalizations alone. Members of both groups had initially indicated their use of the indicator length of stay in hospital due to fall-related injuries as a proxy for injury severity; however, further discussions noted how the effects of existing comorbidities as well as the patient’s discharge status may confound the relationship between injury severity and the length of a hospital stay. This resulted in this indicator ultimately being de-prioritized (Table 2). The need for a more appropriate indicator to capture this construct was reported, and reflected in the high-average ranking of the rate of serious fall related hospitalizations for the PH group.
The importance of severity for both groups is further demonstrated in the discourse around the indicator disability adjusted life years due to a fall. Participants indicated that they had previously used this measure in an attempt to capture the severity of fall-related injuries; however, it was reported that this indicator does not sufficiently reflect the concept of severity for an older adult population over its use in a younger population. This is reflected in the indicator ranking lowest for MS, and de-prioritized outside of the top 10, for PH.
Interestingly, both groups ranked the proportion of older adults with a comprehensive falls risk screening and assessment completed similarly - 4th and 5th for MS and PH, respectively. Those primarily working in healthcare settings in the MS group indicated that this indicator would be beneficial both as a proportion of older adults with a risk assessment and follow-up completed, but also with the ability to disaggregate risk assessments from follow up, along the continuum of care. This indicator could be used to take action to prevent an initial fall by knowing the number of older adults without a completed risk assessment, to secondary prevention of a subsequent fall for those at increased risk. There is a breadth of literature that supports risk assessment and follow up as an effective intervention for older adults.15 Public health also placed this indicator 5th. This could be due to PH efforts to align their role as advocates for increasing reach and adoption of effective interventions in this population.
Another discrepancy between the groups was the ranking of the indicator direct and indirect costs associated with falls. The MS group had this as their second highest ranked indicator, while it did not make it into the top 10 ranked indicators for PH. During discussions regarding this indicator, participants across sectors expressed that being able to present the cost of a high-burden health issue to decision-makers, as is the case with falls, would prompt action and help make a case for investment in prevention programming. For those in PH, participants indicated that presenting the cost of falls often does not actually influence decision-makers in Public Health to prioritize this work in practice, subsequently resulting in it being ranked lower.
Overall, the differing functions of public health compared to other health sectors was clearly reflected in each group’s respective priorities. While both groups prioritized indicators that capture the population burden of fall-related injuries, the PH group was more likely to prioritize indicators that report on the context and severity of falls and would be useful for population-level intervention planning. Conversely, the MS group was more likely to prioritize indicators that measure system performance. The indicators included in the top 10 were similar across both groups (8 the same, 2 different) their respective orders varied considerably, reflecting these different priorities. While foci differ between MS and PH, together the two sets of indicators represent a whole system approach to older adult falls prevention, ranging from primary to tertiary prevention.
Strengths and Limitations
A strength of this work is the systematic and iterative process of gathering input from injury prevention practitioners in Ontario, ensuring that perspectives from each sector were represented at each stage and results being revised accordingly. Additionally, these methods were informed by previous work in this area,11-13 and experts in injury indicator development provided feedback and guidance throughout. The results of this work are novel to Ontario, with many indicators being identified and specified for the first time in the province. Identifying the priority indicators for PH and across sectors in Ontario will help facilitate their systematic use and support consistent reporting over time and across sectors and regions.
There are, however, several limitations to this work. Firstly, the literature search only included English-language sources, and it is possible that indicators from sources published in other languages were missed. Secondly, due to the redeployment of many potential participants for this work to the COVID-19 pandemic response, participation rates were relatively low, with only 19 respondents to the survey. While we achieved participation from all relevant sectors, many only had one representative and therefore may not have accurately reflected the thoughts of others in that sector. Additionally, during some feedback discussions, some sector representatives contributed disproportionately; as a result, we completed a pairwise comparison activity in an effort to ensure equal representation from each sector that ensured equal weighting from each respondent.
Implications for Practice
This work clearly identifies the fall prevention indicator priorities for public health, as well as across health sectors in Ontario. Our list of prioritized indicators align with previous work in this area.16 Oakey et al. (2022) completed a similar process of ranking indicators based on importance and actionability with key decision-makers in British Columbia, Canada.16 Eight indicators were prioritized for fall prevention including the top three related to burden (mortality, hospitalizations and ED visits), direct and indirect costs, wait time for surgery, health service coverage (which includes falls risk screening assessment, strength and balance exercise program, and home risk assessment coverage), dedicated fall prevention staff and the availability of fall prevention resources and plans.16 The fall prevention indicators prioritized by decision-makers in British Columbia echo those prioritized in Ontario and strengthen our understanding of the intelligence required from data of those working in fall prevention across provinces. Identifying and prioritizing these indicators responds to the needs expressed by injury prevention practitioners for indicators that better reflect their work in fall prevention; however, indicators prioritized by this group should be considered within the suite of indicators prioritized across sectors. This accounts for the reality that all systems have specialist areas, and specialist work foci within them; with emphasis that the system works most efficiently and effectively when all acknowledge and work within the system, rather than beside each other in unique (albeit complementary) systems. This will support the systematic use of indicators for reporting on fall-related injuries over time and across geographic regions, and to inform fall-prevention program development, implementation and evaluation.
Further, the differing priorities between sectors highlights the need for indicators that are useful for those working in different health sectors and the importance of consulting practitioners when setting priorities. This work will support movement toward a system that centralizes data and indicator reporting in a consistent and systematic way and fulfills the actual needs of its end users.