Staffing analytics

Measuring Vacancy Duration for Home-Care Staffing Decisions

Vacancy duration becomes useful only after a team defines when a requisition opens, pauses, fills, closes, and remains unresolved.

Staffing Care Home Editorial TeamPublished 10 min read
On this page8 sections

“Time to fill” sounds like a simple difference between two dates. In home care, the start and end points are often ambiguous. A branch may recruit continuously rather than open one requisition for one seat. One posting may cover several service zones and schedules. A candidate may accept an offer but wait for authorized onboarding steps, while client demand changes during the wait. This review offers a cohort method that makes vacancy duration interpretable without claiming that broad labor statistics predict one agency’s hiring timeline.

Research question and public evidence

The research question is: how should a home-care agency define and analyze unresolved staffing demand over time? The BLS Job Openings and Labor Turnover Survey publishes estimates for broad industries. The BLS occupational outlook profile and Occupational Employment and Wage Statistics provide occupation context. Census Quarterly Workforce Indicators provide longitudinal workforce measures by geography and industry.

These sources can frame labor-market conditions and definitions. JOLTS health care and social assistance estimates are broader than home care and are not a direct benchmark for a branch. OEWS is not a vacancy survey. QWI measures do not identify the age of an agency requisition. Local requisition and candidate records must establish vacancy duration.

Methodology

First define the demand unit. For a conventional role, the unit may be an approved requisition with a specified number of positions, branch, service area, schedule class, and minimum criteria. For continuous caregiver recruiting, create demand cohorts at fixed intervals with an approved capacity target and clear schedule or geographic scope. Do not call a general applicant pool one vacancy unless it represents documented hiring demand.

Define the opening date as the date authorized recruiting may begin after required approvals are complete. Preserve earlier request and approval timestamps separately so approval delay can be studied without being mislabeled as recruiting delay. Define a filled position as a candidate reaching the organization’s approved terminal milestone, such as completed start readiness or first eligible assignment. State the milestone exactly.

Requisitions also need pause and closure rules. A pause may stop active recruiting because demand changed or authorization is pending. Decide in advance whether paused days remain in calendar duration and report both gross and active recruiting time when useful. Closure reasons should distinguish filled, demand withdrawn, duplicate, restructured, no longer funded, or administratively closed. An unfilled closure is not a successful fill.

Use a cohort table with requisition identifier, opening date, capacity count, scope, pause intervals, candidate-stage events, positions filled, closure date, and closure reason. At the analysis cutoff, open requisitions are right-censored: their final duration is unknown. Keep them in the analysis instead of deleting them.

Evidence analysis

Begin with a cohort flow showing positions authorized, requisitions opened, still open at cutoff, filled, and closed unfilled. Report the median and selected percentiles only when the cohort is large enough, alongside the full count. An average alone can be distorted by a few very old records.

A survival-style table is often clearer than one time-to-fill number. For each elapsed-time band, show how many requisitions were still under observation at the start, how many achieved the defined fill milestone, how many closed for another reason, and how many reached the cutoff unresolved. This keeps current vacancies visible and prevents completed cases from dominating the story.

Separate recruiting stages to locate delay. Measure opening to first qualified review, review to interview, interview to conditional offer, offer to acceptance, and acceptance to the approved start milestone. These intervals describe process location; they do not prove the responsible cause. A long final interval might involve candidate choice, required documentation, external verification, orientation capacity, or an internal handoff.

Stratify cautiously by service zone, schedule class, requisition type, and recruiting channel. Compare only groups with similar role definitions and observation windows. A rural overnight cohort and a general daytime pool are not interchangeable. Public BLS and Census data may help describe the surrounding market, but the local comparison should not attribute a difference to the market without a stronger design.

Candidate flow as supporting evidence

Vacancy age should be read beside the candidate funnel. Record completed applications, reviews, interviews, offers, acceptances, withdrawals, and completed starts for each demand unit. Deduplicate candidates according to a written rule. Preserve “pending” separately from “failed,” and use specific authorized disposition categories.

The funnel can distinguish low applicant volume from later-stage delay. It cannot show candidate quality merely from a count. Nor should leaders maximize progression by weakening job-related requirements. Requirements must remain approved and consistently applied, with employment decisions made by authorized staff.

Show data completeness in the same report. Missing stage timestamps, generic dispositions, merged requisitions, and retroactively created opening dates can create apparent delays. If data quality improves during the period, annotate the change rather than presenting it as a recruiting performance shift.

Validation and reproducibility

A reviewer should sample recently filled, long-open, paused, canceled, and reopened requisitions. Trace approval, opening, candidate events, fill milestone, and closure to source records. Recalculate active and gross duration. Confirm that a multi-position requisition decrements capacity one position at a time instead of closing at the first start.

Publish a measure specification containing time zone, cutoff date, scope, demand-unit definition, start and end rules, pause treatment, censoring method, reopening treatment, deduplication, and exclusions. Store the extraction query or documented procedure. A future reviewer should be able to recreate every aggregate without relying on undocumented spreadsheet edits.

Operational interpretation

The evidence can support targeted experiments. If most observed delay occurs before first review, leaders can test queue ownership and compare a later cohort. If accepted candidates wait at one handoff, leaders can review that handoff’s capacity and prerequisites. Keep the same milestone definitions and check for concurrent demand changes.

Do not convert the oldest-open list into automatic blame. Old age can signal a difficult schedule, an obsolete requisition, a process bottleneck, or missing closure. Each requires a different response. The report should route exceptions to an authorized owner with the source evidence attached.

Limitations

Local requisition data are observational and shaped by changing client demand, seasonality, geography, role scope, and applicant choice. Continuous recruiting makes vacancy units difficult to define. Small cohorts produce unstable duration estimates. Right-censoring means unresolved cases do not yet have final outcomes, while reopened requisitions complicate elapsed time.

JOLTS, OEWS, occupational outlook, and QWI sources use different populations, periods, and classifications. They cannot provide a direct target for local time to fill. This methodology does not establish hiring quality, discrimination compliance, retention, or service sufficiency. It also does not determine whether a criterion or disposition is lawful; qualified employment reviewers retain that authority.

Conclusion

Vacancy duration becomes credible when the agency defines the demand unit, opening authority, fill milestone, pauses, closures, and unresolved cases before calculation. Cohort flows and stage intervals provide more evidence than a single average. Public labor sources offer context, while local versioned records provide the result. With censoring, validation, and explicit limitations, home-care leaders can locate recruiting delay without pretending that a broad national openings estimate predicts their next caregiver start.

Sources

<- Back to research