Families starting services within 14 days rose from 30% to 70%
At one community-based human-services organization, CARES and the organization’s staff replaced informal referral handoffs with a shared intake workflow, named ownership, and a review process for stalled referrals.
Eligible referrals completing a first service within 14 days
36 of 120 compared with 88 of 126
The result, with its denominator attached.
Comparing eligible referrals received in October–November 2024 with those received in May–June 2025, the share completing a first service within 14 calendar days rose from 36 of 120 to 88 of 126.
- What changed
- A shared register, intake-readiness checklist, accepted handoffs, pooled openings, and weekly review replaced local tracking and unowned forwarding.
- What was measured
- The proportion of eligible new household referrals completing a substantive first program appointment within 14 days of referral receipt.
A referral could be “sent” and still belong to nobody.
Referrals arrived through a shared mailbox, partner forms, and telephone calls. Intake employees recorded information locally before forwarding it to program teams. Families sometimes had to repeat their circumstances because the receiving employee could not determine what had already been collected.
Neither team consistently owned the interval between referral receipt and a completed appointment. A referral could be marked “sent” by intake and remain unassigned in the receiving program. Families experienced repeated calls and uncertain next steps, while supervisors reconstructed referral histories when partners requested updates.
Earlier reminders and revised forms had not solved the problem. They did not establish shared status definitions, a named accepting owner, or a deadline for completing the handoff.
The team defined “started” before comparing performance.
CARES and the organization’s intake supervisor reconciled referral records, contact notes, and completed-visit logs. The rules below were applied to both cohorts.
- Starting services meant completing a substantive program appointment. A screening call or appointment booking did not count.
- Eligible referrals were new household referrals meeting program eligibility. Duplicates, ineligible referrals, and urgent cases handled through a separate response pathway were excluded.
- Families who declined, could not be reached, or remained waiting stayed in the denominator.
- Elapsed time began on the original referral-receipt date. The clock did not pause for missing information, rescheduling, or family-requested delays.
- Every referral had a fixed 45-day follow-up window.
The bottleneck was not simply appointment supply.
Program calendars contained unused intake openings. The assessment instead found three operating conditions: information requirements varied by employee; handoffs lacked an accepting owner and deadline; and the status “waiting” combined materially different situations, including missing information, family preference, and unavailable appointments.
Management chose to standardize the current workflow before considering new software. Program leaders also separated information genuinely required before a first service from information that could appropriately be completed later.
Four controls turned a forwarding habit into an owned workflow.
- 01
One referral register and status dictionary
Every open referral required a named owner, next action, and due date. Role-based access limited the register to the minimum operational information; detailed family narratives stayed in the existing case-record system.
- 02
A common intake-readiness checklist
Program supervisors agreed on the minimum information needed before a first service and documented true program-specific exceptions. The team tested the checklist against previously delayed referrals and revised ambiguous requirements.
- 03
Accepted handoffs with a deadline
Intake remained accountable until a program coordinator accepted the referral. Any handoff still unaccepted after one business day escalated to the program operations lead.
- 04
Pooled openings and a short review routine
Program coordinators pooled existing first-appointment openings. A 15-minute weekly review addressed overdue actions and appointment constraints. CARES coached the initial meetings, then the intake supervisor led them.
Two intake employees and two program coordinators completed scenario-based training. A two-week pilot covering 18 referrals exposed duplicate entry, so the team removed a parallel tracking sheet before expanding the workflow. The organization added no intake positions or service hours.
The improvement appeared across more than one measure.
Scroll horizontally or use the arrow keys to compare all measures.
| Measure | Oct–Nov 2024 | May–Jun 2025 | Evidence basis |
|---|---|---|---|
| First service completed within 14 calendar days | 36 of 120 · 30% | 88 of 126 · 70% | Reconciled referral records and completed-visit logs |
| First service completed within 45 calendar days | 84 of 120 · 70% | 111 of 126 · 88% | The same records and a fixed 45-day follow-up window |
| Median receipt-to-first-service time among referrals starting within 45 days | 23 days · n=84 | 11 days · n=111 | Date-difference calculation for referrals that started within 45 days |
| No completed first service by day 45 | 36 of 120 · 30% | 15 of 126 · 12% | Day-45 disposition review |
The median includes only families who began services within 45 days, so the completion rates for all eligible referrals remain the principal measures. Among the 36 baseline referrals without a first service by day 45, 12 families declined, 14 could not be reached after the contact protocol, and 10 remained pending. Among the 15 comparison referrals without a first service, six declined, five could not be reached, and four remained pending.
Employees reported fewer searches through email threads and clearer accountability when families called. Supervisors observed more consistent acceptance of handoffs. These observations were not converted into estimated labor savings. Partner confidence and longer-term family engagement were not measured.
The test was whether the workflow continued without CARES.
At handoff, the intake supervisor became process owner and the program operations lead became escalation owner. Two trained employees provided intake backup. The handover package included the register specification, status dictionary, intake-readiness checklist, responsibility matrix, operating procedure, training scenarios, and metric-calculation guide.
Weekly aging reviews continued, supported by a monthly reconciliation of five referral records against appointment logs. Changes to intake prerequisites required program-leader approval rather than informal additions by individual employees.
Absence exercise
15 of 16During a later five-business-day absence exercise, backup employees handled 16 new referrals without CARES routing the work or updating records. Fifteen received a named owner and dated next action by the next business day. One missed the deadline and was corrected through the escalation procedure.
The exercise demonstrated transfer of the intake controls. It did not establish that the earlier 70% service-start rate persisted. Automated appointment-log reconciliation remained unfinished, so management retained manual checks until an integration could be validated.
A queue became a record of ownership and next action.
The representative structure below shows the operating logic without reproducing a client record or family information.
The intake owner remained accountable until acceptance was recorded. Moving a referral into another team’s queue did not, by itself, complete the handoff. Staff described the practical shift as moving from proof that a referral had been forwarded to proof of who accepted it and what the family was waiting for.