Administrative Governance as a Missing Link in Continuity of Care: Policy Recommendations for Modernizing Home Health Services

Full text

Text extracted from the article PDF for reading and search. Tables, figures and equations are in the PDF, which remains the authoritative version.

1. Introduction

A patient discharged from a hospital into Medicare-certified home health care crosses more organizational boundaries in the first 48 hours than almost any other point in the care continuum: the discharging hospital, the certifying physician or allowed practitioner, the home health agency's intake and scheduling functions, the pharmacy, often a durable medical equipment supplier, and frequently a family caregiver who has received no formal handoff at all. Continuity of care is usually reported as a downstream outcome, a readmission rate, a start-of-care delay, a patient experience score, without much visibility into the administrative steps that produced it. This paper treats that visibility gap as the central problem. Referral completeness, eligibility verification, physician order authentication, medication reconciliation, scheduling, and escalation ownership are administrative functions, not clinical ones in the narrow sense, yet failures in any of them can be as consequential for continuity of care as a missed diagnosis.

This critical policy review develops an administrative governance account of continuity of home health care and asks what evidence a defensible policy recommendation requires before it is generalized across agencies, regions, and payer arrangements. Literature coverage extends through 31 December 2024, and the review distinguishes throughout between a proposed governance framework, which this paper supplies, and a validated intervention, which it does not, because no authorized operational dataset was collected or analyzed for this manuscript.

Four questions organize the analysis. First, which administrative functions most plausibly govern continuity of home health care, and how should each be defined and measured against an auditable record rather than a self-reported impression? Second, through what mechanisms does administrative reliability in one function, for example order authentication, propagate into or block continuity in a downstream function, for example the first home visit? Third, which contextual conditions, rurality, language access, payer type, agency size, change the strength of these mechanisms? Fourth, what pattern of evidence, across which comparisons and holdouts, would justify moving a governance recommendation from a pilot into a system-level policy? The analysis adopts a socio-technical view of home health operations. An intake coordinator's workload, an electronic health record's interoperability with the referring hospital, a state's Medicaid prior-authorization rules, and a caregiver's phone access all bear on whether a discharge order becomes a completed home visit. A framework limited to a single technical measure, such as an interoperability score, would misclassify performance by ignoring the organizational and regulatory context in which that measure operates. The governance framework developed in Section 3 therefore integrates outcome, process, context, implementation, and accountability evidence rather than collapsing administrative reliability into one number.

Four evidentiary standards guide the review. Construct validity requires that a proposed measure, for example "order authentication timeliness," represent the concept the Medicare Conditions of Participation and clinical practice actually intend. Internal validity requires that alternative explanations, particularly reverse causation between patient acuity and administrative delay, receive explicit attention before a mechanism is asserted. External validity requires transparent statement of the settings, payer mixes, and time periods to which a claim is meant to generalize. Decision validity asks whether the resulting evidence would actually change a governance choice, such as whether to invest in a new escalation protocol, at the time decision-makers need it. Home health is also, in scale terms, not a peripheral part of the Medicare program: it is among the fastest-growing post-acute benefit categories as the Medicare-eligible population ages and as hospitals face financial and capacity pressure to shorten inpatient stays and shift recovery into the home. That growth raises the practical stakes of the visibility gap this paper addresses. An administrative failure that delays or fragments a single home health episode is a patient-level harm; the same failure, replicated across an agency's caseload or across a regional or national policy rollout, becomes a systemic one, and the Conditions of Participation, the Home Health Quality Reporting Program, and the Home Health Value-Based Purchasing Model all now operate at that scale. A governance framework aimed at policy recommendation, rather than single-agency process improvement alone, has to be built with that scale in mind from the outset, which is part of why this paper treats generalizability conditions, payer type, rurality, agency ownership structure, as a first-class analytic concern in Section 3.4 rather than a caveat appended at the end.

The paper's contribution is deliberately operational. Section 2 synthesizes the relevant literature critically rather than as a list, comparing what different bodies of transitional-care, home health nursing, and readmission research actually established and where they conflict or leave gaps. Section 3 proposes the five-domain governance framework and a corresponding measurement matrix. Section 4 states seven falsifiable propositions with the comparison, adjustment, and holdout conditions a credible test requires. Section 5 specifies a five-stage adoption pathway appropriate to a regulated, accredited industry. Section 6 discusses implications and limitations, and Section 7 concludes. Throughout, novelty, methodological rigor, and fit to the target evidence base matter more than publication tier; a framework paper earns its place in a Q1 outlet by specifying testable claims and minimum evidentiary requirements, not by asserting that adoption is already justified.

2. Critical Synthesis of The Literature

2.1. Foundations: What Continuity of Care Is, And Why

Administration Governs It Continuity of care is not one construct. Haggerty et al. (2003) distinguish informational, management, and relational continuity in their multidisciplinary review, a distinction this paper treats as load-bearing rather than decorative: informational continuity depends on whether a discharge summary, medication list, and plan of care actually reach the next clinician; management continuity depends on whether a coherent, responsive plan of care is maintained across providers; relational continuity depends on an ongoing therapeutic relationship, which is harder to sustain in home health, where staffing turnover and productivity-based scheduling can rotate clinicians across a caseload. Administrative reliability, the argument advanced here, is the load-bearing infrastructure for the first two dimensions and a precondition for the third: a caregiver cannot build a relationship with a clinician who never receives the referral. Donabedian's (1988) structure-process-outcome model supplies the paper's second anchor. Structure (staffing ratios, credentialing, electronic health record capability), process (the sequence of verification, authentication, and reconciliation steps examined here), and outcome (continuity, readmission, patient experience) are causally ordered but not always measured in that order; much of the home health quality literature reports outcomes without the intervening process measures needed to explain them, a gap the Institute of Medicine's (2001) Crossing the Quality Chasm report identified as a system-level weakness two decades ago and that the administrative governance literature reviewed below has still not closed.

2.2. Transitional-Care Intervention Models and What

They Imply For Administration Three intervention models dominate the transitional-care literature and each encodes an implicit theory of which administrative function matters most. The Care Transitions Intervention (Coleman et al., 2006) is built around four pillars, medication self-management, a patient-held record, a scheduled follow-up appointment, and patient knowledge of "red flag" symptoms, and is coach-mediated rather than clinician-delivered; its administrative burden is concentrated in patient activation rather than interagency data transfer. The Transitional Care Model (Naylor et al., 2004; Naylor, 2002) instead places an advanced-practice nurse at the center of a comprehensive, multi-week discharge and home follow-up process for high-risk older adults with chronic illness, shifting administrative load toward continuous case management rather than a single handoff event. Project RED (Jack et al., 2009) targets the hospital side of the transition directly, specifying twelve discrete discharge-process components, several of which, medication reconciliation, a written discharge plan, and confirmed follow-up, are precisely the administrative artifacts a home health agency depends on receiving intact. Kripalani et al. (2007) supply the diagnostic complement to these three intervention studies: their review of deficits in communication and information transfer at hospital discharge documents how often the artifacts these models rely on, discharge summaries, medication lists, pending test results, either arrive late or do not arrive at all. Read together, these four sources do not converge on a single best intervention; they converge on the observation that every credible transitional-care model presupposes an administrative substrate, complete referrals, authenticated orders, reconciled medications, that home health agencies do not uniformly control and that the literature rarely measures directly.

2.3. Hospital-To-Home and Post-Acute Transition

Research A second cluster of studies examines hospital-to-home and post-acute transitions empirically, and its heterogeneity is itself informative. Gill et al. (2010) approach transitional aged care from the patient's own account of quality, a relational-continuity lens; Isac and Lee (2023, 2024) instead synthesize chronic-illness transitional care through integrative and qualitative reviews that foreground unmet informational needs at discharge; Hung and Leidig (2015) evaluate a specific transitional care program's association with readmission reduction among older adults, a management-continuity and outcome focus; and Cerantola and Happ (2012) narrow the lens further to communication-impaired patients moving from intensive care to home, a population for whom standard handoff assumptions, that the patient can confirm receipt of instructions, do not hold. Pincus (2013) and Van der Vlegel-Brouwer (2017) each evaluate transitional care programs for frail older adults but differ in what they treat as the active ingredient, medication reconciliation optimization in the former, program-level effectiveness for the "oldest old" in the latter, while Williams, Rose, and Perez (2017) reframe the same population through an equity lens, asking which vulnerable older adults transitional care programs are actually reaching. Ganann and colleagues (2018, 2021) contribute both a methodological argument, that transitional care research should engage older adults as partners rather than only as subjects, and a pragmatic effectiveness-implementation trial for patients with multimorbidity and depressive symptoms, a design choice that treats implementation fidelity as inseparable from effectiveness. Distelhorst and Hansen (2022) complicate the picture further by showing that neighborhood-level social context, not agency performance alone, predicts heart failure readmission after transitional care, an explicit warning against attributing continuity outcomes to administrative process without controlling for where a patient lives. Chuansangeam (2023) and Joo and Liu (2023) each synthesize the broader hospital-to-home transitional care evidence base, the latter through an umbrella review of interventions for frail older adults, and both converge on a conclusion this paper treats as central: the intervention literature is heterogeneous enough in population, setting, and outcome definition that pooled effect estimates without attention to mechanism are not yet defensible. Boling and Parsons (2007) close this cluster with an explicitly administrative framing, a research and policy agenda for nursing-home-to-home transitions that anticipates several of the governance gaps this paper formalizes. Two conclusions follow from this cluster. First, no single mechanism, medication reconciliation, discharge summary transfer, or follow-up scheduling, is consistently identified as the dominant driver of continuity across these studies; the more defensible reading is that several administrative mechanisms operate jointly and that their relative importance shifts with population and setting, which is the premise on which Section 3's multi-domain framework is built. Second, equity-relevant variation, by neighborhood, by communication ability, by vulnerability status, recurs across this literature closely enough that an administrative governance framework omitting equity monitoring would misrepresent the evidence it claims to build on.

2.4. The Clinical Scope of Home Health Nursing and Why

Administration Cannot Be One-Size-Fits-All A separate, older literature documents how heterogeneous home health caseloads have long been in practice, and this heterogeneity is a direct argument against treating administrative governance as a single uniform checklist. Gerace (1990) and Clark (1991) each describe home health practice for populations, psychiatric home care and HIV management respectively, whose clinical risk profiles bear little resemblance to a typical post-surgical or heart-failure caseload, yet both depend on the same intake, authorization, and scheduling infrastructure. Thobaben (2012), writing two decades later, shows that psychiatric home health skilled nursing remained administratively distinct enough from general home health to warrant its own specialty literature. Casey (1997) documents the emergence of home health nursing specialties under managed care, precisely the period in which authorization and eligibility-verification burdens on agencies intensified, and Marelli (1996) and McGihon (2002) each treat home health nursing practice and its clinical education as requiring distinct competencies from hospital-based nursing, including competencies in the administrative and regulatory environment specific to the home setting. Miller and Daiey (1996) make this distinction explicit in their title, arguing that home health care nursing is not a scaled-down version of hospital nursing, and Biteman (2006) extends the argument to a holistic model of home health practice. Heeke, Wood, and Schuck (2014) and Magilvy and Lakomy (1992) bracket this literature at its more recent and more historical ends respectively, both examining how older adults specifically experience the transition into home care, a population whose administrative needs, medication counts, caregiver availability, transportation, differ systematically from younger home health populations. Indexing entries in Nursing Research spanning the 1970s catalogued community health, home care, and health maintenance organization services as a distinct professional and organizational concern (Nursing Research, 1973, 1975, 1976, 1977, 1978), indicating that administrative and organizational dimensions of home-based care have drawn sustained professional attention for more than five decades, well before the current Medicare Conditions of Participation and OASIS assessment infrastructure existed. Taken together, this cluster supports a specific design choice in Section 3: administrative governance components should be defined so that agencies can apply them across a heterogeneous caseload, psychiatric, infectious disease, cardiac, post-surgical, hospice-adjacent, without assuming a single risk profile.

2.5. Readmission Risk and The Limits of A Single

Administrative Predictor Readmission is the outcome most frequently linked to home health administrative performance, and the literature here is large enough to support a genuine comparative reading rather than a citation list. Several studies isolate a single administrative or clinical lever: Upakdee and Mankongdee (2012) examine pharmacist telephone follow-up for diabetes patients; Kellogg, Brickner, Conley, and Conroy (1991) examine structured control of elderly readmission from a community-based program; Navarro, Enguídanos, and Wilber (2012) develop a claims-based approach to identifying readmission risk among Medicare-aged patients; and Cole, Lush, Dawson-Rose, and Waters (2020) examine factors influencing 30-day readmission specifically during the home health service window, closer to this paper's own outcome boundary than most of the cluster. A second group foregrounds communication rather than a single clinical lever: Jones and Levy (2019) argue directly that improved communication in home health could reduce hospital readmission rates, and Pesko, Gerber, Peng, and Press (2017) examine nurse-physician communication and patient severity as joint predictors of readmission, a design that treats communication quality as measurable rather than assumed. A third group is condition-specific: Martens (2000), Proctor, Morrow-Howell, Li, and Dore (2000), Li, Morrow-Howell, and Proctor (2004), and Anderson, Clarke, Helms, and Foreman (2005) each examine home care and readmission specifically for congestive heart failure patients across different eras and payment environments, including the shift to prospective payment, while Lodge, Bashiri, Durazo-Arvizu, and Russell (2024) examine an entirely different population, pediatric tracheostomy patients, and Treece et al. (2018) examine hospice rather than skilled home health. Damiani et al. (2009) contribute an Italian health-system comparison of discharge planning models, and Hansen, Williams, and Singer (2010) link hospital safety climate perceptions directly to readmission incidence, a structure-level predictor distinct from any single discharge process step. Nuckols (2015) and Dimick and Miller (2015) each caution, from different angles, county-level policy variation and surgical readmission respectively, against treating readmission reduction programs as uniformly effective; both argue that readmission is influenced by factors substantially outside any single agency's administrative control, including community resources and case mix. Cunha Ferré et al. (2019) narrow the outcome window further, to 72-hour readmission after discharge with home care services in an Argentine cohort, a granularity most of the cluster does not attempt. The comparative lesson is not that one predictor dominates but that population (CHF versus pediatric versus general elderly), setting (domestic United States versus Italy versus Argentina), and outcome window (72 hours versus 30 days) vary enough across this literature that pooling these studies into a single readmission-prediction claim would be a methodological error. What is consistent across the cluster is the recurring implication that communication and information transfer, not a single clinical variable, repeatedly surface as either the mechanism under study or the confound investigators had to control for, which is the empirical basis for placing handoff and interprofessional communication at the center of the governance framework in Section 3 rather than treating readmission risk scoring as a self-sufficient solution.

2.6. Medication Reconciliation: Consistent Process Gains, Inconsistent Utilization Effects The medication reconciliation literature offers the clearest example in this review of a mechanism with strong process-level evidence and weaker utilization-level evidence, a distinction a governance framework must preserve rather than collapse. Cochrane systematic reviews of medication reconciliation interventions across transitions of care report that structured interventions reliably reduce medication discrepancies but yield inconsistent, generally low-certainty evidence of effect on hospital utilization outcomes such as readmission or emergency department visits (Redmond et al., 2013, 2018). This distinction matters directly for Section 3: it argues for measuring medication reconciliation as a process indicator in its own right rather than assuming its value is fully captured by a downstream readmission association. Consistent with that reading, Becerra (2019), Salman and Vo (2020), and Bartzak (2022) each document how incomplete reconciliation during transitions of care creates identifiable safety and documentation gaps without always tracing those gaps to a readmission outcome, while Burch and Ronellenfitsch (2019) frame the same evidence base as a live open question, what effects, precisely, medication reconciliation interventions reliably produce. Martin (2012) and Liu and Garwood (2015) both approach reconciliation from a pharmacy-practice error-avoidance perspective, and Oraibi, Holmes, and Pichardo-Lowden (2020) and Heffner, Dillaman, and Hill (2020) each demonstrate condition-specific reconciliation gains, diabetes medication optimization and oral oncolytic error reduction respectively, in pharmacist-driven models, suggesting the mechanism generalizes across therapeutic areas even where downstream utilization effects remain uncertain. Roure Nuez et al. (2016) and Tamiru, Edessa, Sisay, and Mengistu (2018) extend the evidence internationally, an electronic clinical decision support tool in Spain and discrepancy rates in an Ethiopian tertiary hospital respectively, while the AMCP Partnership Forum (2014) reports a policy-level consensus process on electronic solutions to medication reconciliation, closer to the regulatory and technology-infrastructure layer this paper's governance framework also addresses.

2.7. Regulatory Anchors and An Access-Equity

Complement The review's regulatory anchors, the Medicare home health Conditions of Participation (Centers for Medicare and Medicaid Services, 2024), Agency for Healthcare Research and Quality patient safety guidance on care transitions (2023), and the American Case Management Association's transitions-of-care standards (2023), do not themselves report empirical findings but define the compliance floor against which several of the administrative components in Section 3 are specified, including the face-to-face encounter requirement, OASIS-anchored assessment, and quality assessment and performance improvement obligations. Finally, work on artificial-intelligence-enabled diagnostics and scalable access models (Oluwasegun, 2024) is not home health-specific but is relevant to one recurring equity concern in this review: administrative reliability measures such as scheduling timeliness and language access are likely to interact with the same rural and underserved-community access barriers that AI-enabled point-of-care diagnostics research is attempting to address elsewhere in the health system, which is why equity monitoring appears as a governance domain in Section 3 rather than as an afterthought.

2.8. What the Synthesis Establishes and What It Leaves

Open Three conclusions follow from this synthesis and organize the rest of the paper. First, no single study or model cluster identifies one administrative mechanism as sufficient for continuity of home health care; the literature instead supports a small set of jointly necessary mechanisms, referral and eligibility work, order authentication, medication reconciliation, handoff communication, and escalation, that this paper organizes into governance domains. Second, methodological heterogeneity, in population, setting, outcome window, and country, means that pooled effect estimates are premature; a credible research agenda must first standardize definitions and measurement before attempting synthesis across studies. Third, publication bias and selective outcome reporting remain a live risk across this literature; several of the readmission and reconciliation studies reviewed above report only positive associations, and protocol registration, complete measure dictionaries, and reporting of null results would materially strengthen the evidence base this paper draws on.

2.9. From Synthesis to Policy Relevance

Three features of this literature carry directly into the policy question this paper is ultimately organized around, whether and how administrative governance recommendations should be generalized beyond a single agency. First, the intervention literature in Section 2.2 shows that credible transitional-care models already presuppose administrative reliability as a precondition rather than treating it as incidental; a policy that funds care-transition coaching or follow-up scheduling without also addressing the referral, order-authentication, and reconciliation mechanisms documented in Sections 2.3 and 2.6 is funding only part of the causal chain those models depend on. Second, the population and setting heterogeneity documented across Sections 2.3 through 2.6, CHF-specific versus general elderly versus pediatric caseloads, urban versus rural settings, domestic versus international health systems, means that a single national administrative-reliability threshold is unlikely to be appropriate; a policy recommendation defensible for a large, multi-site urban agency may not transfer, without the holdout testing specified in Section 6, to a small rural or hospital-based agency operating under different referral patterns. Third, the equity-relevant variation recurring across Section 2.3 in particular is not a peripheral finding to be addressed after a governance framework is otherwise finalized; it is evidence that any national or state-level policy built on this framework should require equity-stratified reporting as a condition of adoption, not as a voluntary enhancement, because the studies reviewed here consistently find that averaged, unstratified outcome reporting can obscure exactly the disparities policymakers most need to see.

3. An Administrative Governance Framework for

Continuity of Home Health Care

3.1. Five Governance Domains

Rather than treat administrative governance as an undifferentiated category, this paper organizes it into five domains, each with a distinct set of mechanisms, a regulatory or practice anchor, and a primary reliability measure. The domains are ordered to reflect the typical sequence of a home health episode, though feedback across domains is expected and is itself part of governance.

Referral and intake governance covers eligibility verification, homebound-status and skilled-need documentation, and insurance or Medicare Advantage authorization. Its regulatory anchor is the Conditions of Participation's intake and eligibility requirements together with the American Case Management Association's transitions-of-care standards (ACMA, 2023). Failures here are often invisible in outcome data because they occur before a case is opened, an episode that never starts leaves no readmission to measure. The clinical-administrative interface covers physician order authentication under the face-to-face encounter requirement and OASIS-anchored plan-of-care development. Its anchor is the Affordable Care Act's face-to-face provision as implemented in current Conditions of Participation and the OASIS data set (CMS, 2024). This domain sits at the boundary between clinical judgment and administrative documentation, and it is where an unauthenticated verbal order can stall an otherwise ready case. Care-transition and handoff governance covers receipt of the hospital discharge summary, medication reconciliation against that summary, and structured interprofessional communication at the first visit. Its conceptual anchor is the transitional-care and information-transfer literature reviewed in Section 2, in particular Kripalani et al. (2007), Coleman et al. (2006), and Jack et al. (2009). This is the domain most directly implicated by the medication reconciliation and communication findings summarized above. Escalation and safety-net governance covers after-hours coverage, change-in-condition response, missed-visit recovery, and risk stratification. Its anchor is the Conditions of Participation's care planning and quality assessment and performance improvement (QAPI) requirements. This domain is where a home health agency's reliability is tested outside business hours, when a caregiver notices a symptom change and needs to reach someone who can act. Quality and compliance governance covers Home Health Quality Reporting Program measure submission, Home Health Value-Based Purchasing performance review, incident and complaint resolution, and accreditation survey closure under CHAP, ACHC, or Joint Commission standards. This domain closes the loop: it is where an agency learns whether the first four domains are functioning and decides whether to retain, adapt, or escalate a governance change. Table 1 summarizes the five domains, their core mechanisms, their regulatory or practice anchor, and the primary reliability measure this paper proposes for each.

Table 1: Administrative governance domains in Medicare-certified home health Domain Core administrative mechanisms Regulatory or practice anchor Primary reliability measure Referral and intake governance Eligibility verification; homebound and skilled-need documentation; insurance or Medicare Advantage authorization Conditions of Participation intake requirements; ACMA transitions-of-care standards (ACMA, 2023)

Percentage of referrals with all required fields complete at eligibility determination; referral-to-eligibility-determination time Clinical-interface Physician order authentication; face-to-face encounter documentation; OASIS-anchored plan of care Face-to-face encounter requirement; OASIS data set (CMS, 2024)

Percentage of verbal orders authenticated within the required window; percentage of episodes with complete face-to-face documentation before claim submission Care-transition and handoff governance Discharge summary receipt; medication reconciliation; structured interprofessional handoff at first visit Transitional-care and information-transfer literature (Kripalani et al., 2007; Coleman et al., 2006; Jack et al., 2009)

Percentage of start-of-care visits with a discharge summary and a reconciled medication list available Escalation and safety-net governance After-hours coverage; change-in-condition response; missed-visit recovery; risk stratification Conditions of Participation care planning and QAPI requirements Time from reported change in condition to clinician response; missed-visit recovery rate within 48 hours Quality and compliance governance HHQRP measure reporting; HHVBP performance review; incident and complaint resolution; accreditation survey closure CMS Home Health Quality Reporting Program and Value-Based Purchasing Model; CHAP, ACHC, or Joint Commission accreditation standards Audit-finding closure time; trend in readmission and emergency-department-use measures

3.2. How the Domains Interact

The five domains are not independent. A referral that arrives incomplete (domain 1) increases the probability that a verbal order will be taken and not promptly authenticated (domain 2), which in turn increases the probability that the first visit occurs before medication reconciliation is complete (domain 3). A missed first visit (domain 3) raises the probability that an after-hours call will be the patient's first contact with the agency (domain 4), which raises the load on escalation staff and, if repeated, will surface in incident reports and Home Health Value-Based Purchasing composite performance (domain 5). Governance, in this framework, is the set of decision rights, review cadences, and escalation rules that determine who notices this chain and who is authorized to interrupt it. A domain-level measure examined in isolation, an order-authentication timeliness rate with no reference to referral completeness, will therefore understate or overstate an agency's actual continuity risk depending on where in the chain the underlying failure originates. This is the paper's central argument against single-score quality dashboards: a composite that averages across domains can mask a bottleneck in exactly the domain where intervention would be most effective.

A concrete illustration makes the chain easier to test against real operations. Consider a referral that reaches an agency on a Friday afternoon for a patient discharged after cardiac surgery, with an incomplete medication list and no confirmed physician signature on the plan of care. Under ordinary staffing, eligibility verification (domain 1) may not complete until Monday, which pushes order authentication (domain 2) into the same window as the first visit rather than ahead of it. If the first visit proceeds regardless, because CoP timeliness requirements press against the delay, the clinician arrives without a reconciled medication list (domain 3) and must either delay reconciliation or reconcile against an incomplete source, either of which raises the probability that a medication-related change in condition will be reported over the weekend, precisely when escalation coverage (domain 4) is thinnest. Whether that call is answered by a clinician with authority to adjust the medication regimen or is triaged to a general answering service is a governance decision, not a clinical one, and it is the kind of decision Table 3's implementation-fidelity standard is designed to make visible after the fact through incident review (domain 5). None of the individual domain measures in Table 1, examined alone, would necessarily flag this chain as a single governance failure; only cross-domain review would.

3.3. Governance Decision Rights and The Limits of

Automation Each domain requires an explicit answer to four governance questions: who owns the data produced in that domain, who reviews it and on what cadence, who has authority to change a classification, escalate a case, or pause a workflow, and how an affected patient, family caregiver, or staff member corrects an error. Electronic health record and referral management systems can support these functions, timestamping a referral, flagging an unauthenticated order, but they cannot substitute for the governance decision about who is accountable when the system flags a problem. Interoperability between hospital and home health electronic records remains uneven; home health agencies were largely excluded from the meaningful-use incentive programs that accelerated hospital and physician electronic health record adoption, and many still receive discharge information by fax, secure email attachment, or courier rather than through direct system-to-system exchange, even where the discharging hospital's own record is fully electronic. A governance framework built on the assumption of seamless data exchange would therefore misdescribe the operating reality of most Medicare-certified agencies, and the discharge-summary-receipt measure in Table 2 is defined broadly enough, received before first visit, regardless of transmission method, to remain valid whether an agency's receiving workflow is a certified interoperability interface or a staff member checking a fax queue. Automation should therefore enter a domain only after its definitions, workflow, and escalation ownership are stable under manual or semi-manual operation, consistent with the phased adoption pathway developed in Section 5.

3.4. Boundary Conditions and Scope

The framework is scoped to Medicare-certified home health agencies operating under 42 CFR Part 484 and is not intended to generalize without adaptation to private-duty home care, hospice, or non-certified personal care services, which operate under different certification, staffing, and payment rules. Within Medicare-certified agencies, the framework should still be expected to perform differently across at least three boundary conditions. First, traditional Medicare and Medicare Advantage referrals carry different authorization burdens; Medicare Advantage plans frequently require prior authorization for home health episodes that traditional Medicare does not, which changes the referral-and-intake domain's baseline timeline independent of any agency-level governance change. Second, state licensure and Medicaid managed care rules vary enough that a governance change validated in one state should not be assumed to transfer to another without the held-out comparison specified in Section 6. Third, agency size and ownership structure, an independent single-site agency, a multi-site regional chain, or a hospital-based agency, changes which entity plausibly owns each domain's decision rights in Section 3.3; a hospital-based agency, for example, may have direct electronic health record access to the discharging hospital that a freestanding agency does not, which would be expected to strengthen the care-transition and handoff domain's baseline performance independent of any governance intervention. A defensible test of the propositions in Section 5 should report which of these boundary conditions applied to the setting studied.

4. Component Measurement and Reliability Indicators

Within each governance domain, Section 3 identifies mechanisms that require their own operational definition before they can support a policy decision. Table 2 specifies twelve such components across the five domains, each with its construct role, the evidence source an agency would use to measure it, and the timing at which measurement is most informative. The list is deliberately smaller than an exhaustive enumeration of every administrative touchpoint in a home health episode; components were retained where the literature reviewed in Section 2 gives a specific reason to expect they matter for continuity, and dropped where inclusion would have been definitional padding rather than an evidence-linked claim.

Table 2: Component measurement matrix by governance domain Component Domain Construct role Primary evidence source Measurement timing Referral completeness Referral and intake Process indicator Referral record audited against a required-field checklist At eligibility determination; repeated at plan-of-care review Eligibility and homebound documentation Referral and intake Process and compliance indicator Intake record compared against CoP homebound criteria At intake; repeated at recertification Insurance or Medicare Advantage authorization Referral and intake contextual moderator Authorization log and payer correspondence At intake and at each authorization renewal Order authentication timeliness Clinical-interface Process and compliance indicator Physician order log; signed-order turnaround time Continuous; reviewed weekly Face-to-face encounter documentation Clinical-interface Compliance indicator F2F encounter note checked against the CoP window At certification and recertification OASIS-anchored plan-of-care completeness Clinical-interface construct-validity anchor OASIS assessment compared against the plan of care Start of care, resumption of care, recertification Discharge summary receipt and completeness and handoff Process indicator Hospital discharge summary checked against required elements Before or at first visit Medication reconciliation accuracy and handoff Outcome-adjacent process indicator Reconciled medication list compared against hospital and pharmacy source lists At first visit; repeated at each subsequent transition Structured handoff completion and handoff Process and fidelity indicator Handoff documentation audit (for example, an SBAR-formatted note)

At first visit and at any clinician change After-hours escalation response time Escalation and safety-net outcome-adjacent On-call log; response timestamp Continuous; reviewed monthly Missed-visit recovery rate Escalation and safety-net Process indicator Visit schedule compared against the completed-visit log Continuous; reviewed weekly Equity-stratified reliability gap Cross-cutting, all domains Contextual and equity indicator Stratified analysis by rurality, language, and payer Quarterly, aligned with QAPI review cadence

4.1. Referral and Intake Components

Referral completeness and eligibility documentation are best measured at the point eligibility determination begins, not after intake staff have already filled gaps by phone, because the gap-filling effort itself is a workload signal worth tracking separately from the referral's original quality. A referral missing a face sheet, a recent history and physical, or a clear physician signature forces intake staff into a reactive callback cycle that delays eligibility determination independent of clinical acuity. Insurance and Medicare Advantage authorization sits in this domain because, unlike traditional Medicare, several Medicare Advantage plans require prior authorization before a home health episode can begin, and an agency's authorization turnaround time is therefore as much a payer-relationship measure as an internal process measure.

4.2. Clinical-Administrative Interface Components

Order authentication timeliness should be measured from the moment a verbal order is taken to the moment a physician or allowed practitioner's signature is on file, and agencies should distinguish routine recertification orders from urgent plan-of-care changes, since the two carry different clinical risk if delayed. Face-to-face encounter documentation is a compliance measure with direct financial consequence, an incomplete F2F record can result in a claim denial independent of the quality of care delivered, which makes it one of the few components in this framework where the administrative and financial stakes are explicit rather than inferred. OASIS-anchored plan-of-care completeness should be checked for internal consistency, whether the functional and clinical status recorded in the OASIS assessment is reflected in the goals and interventions in the plan of care, since a mismatch between the two is itself evidence of a process breakdown rather than a data-entry error.

4.3. Care-Transition and Handoff Components

Discharge summary receipt should be measured as a binary indicator (received before first visit, yes or no) and as a completeness score against a defined element list, medication list, pending results, follow-up appointments, because a summary that arrives but omits pending laboratory results creates a distinct and separately measurable risk from a summary that never arrives at all. Medication reconciliation accuracy should be measured against both the hospital discharge medication list and, where available, the pharmacy fill record, since the two sources do not always agree and an agency that reconciles only against the discharge summary may miss a pharmacy-level discrepancy. Structured handoff completion should record not only whether a handoff occurred but which structured elements, situation, background, assessment, and recommendation in an SBAR-formatted note, were actually completed, since a checkbox marked "handoff complete" with no substantive content is a fidelity failure this framework treats as distinct from no handoff at all.

4.4. Escalation and Safety-Net Components

After-hours escalation response time should be measured from the moment a call or alert is logged to the moment a clinician with authority to act responds, and agencies should record whether the response was by phone, telehealth, or an in-person visit, since the appropriate response depends on what the caregiver is reporting. Missed-visit recovery rate should be measured as the percentage of missed visits rescheduled and completed within 48 hours, a window chosen because clinical deterioration risk after a missed skilled visit rises measurably within that period for many home health diagnoses. Risk stratification should be validated against actual after-hours escalation and unplanned hospitalization outcomes within the agency's own caseload rather than imported unchanged from a different population, since Section 2.5 documents substantial population heterogeneity, CHF, pediatric, hospice, general elderly, across the readmission literature this framework draws on.

4.5. Cross-Cutting Equity Monitoring

The equity-stratified reliability gap is deliberately placed outside the five sequential domains because it applies to all of them. Agencies should compute each Table 2 measure separately by rurality, preferred language, and payer type at least quarterly, aligned with the QAPI review cadence, and should treat a widening gap in any domain as a governance finding requiring the same response as a missed compliance deadline, not as a secondary observation to revisit later. Section 2.3's synthesis of vulnerability-focused transitional-care research (Williams et al., 2017; Distelhorst and Hansen, 2022) gives specific reason to expect that averages computed without this stratification will understate the reliability gap experienced by rural and limited-English-proficiency patients.

Three measurement principles apply across all twelve components. First, each measure needs an explicit denominator and observation window, referral completeness, for example, should be measured as a percentage of referrals with all required fields present at the point eligibility determination begins, not at some later, already-corrected state, because correction timing is itself an administrative-burden signal worth tracking separately. Second, missingness should be reported alongside the measure itself; a component with thirty percent missing data is not simply performing at whatever rate the complete cases show, and the framework treats missingness as a governance finding in its own right, often indicating that the workflow generating the underlying record is not yet stable enough for the measure to be trusted. Third, every component should be stratified at minimum by referral source, payer type, rurality, and preferred language, because the equity-relevant variation documented across the transitional-care literature in Section 2.3 gives specific reason to expect that administrative reliability is not uniform across these strata, and a framework that reports only agency-level averages would obscure exactly the disparities that literature identifies.

4.6. Evaluation and Validation Standards

Measuring the twelve Table 2 components is necessary but not sufficient; the framework also needs an explicit standard for judging whether a given measure is trustworthy and decision-relevant before it is used to justify a governance change. Table 3 states eight standards and, for each, the specific interpretation question an agency or evaluator applying this framework to home health data would need to answer, rather than a generic methodological checklist that could apply to any measurement problem.

Table 3: Evaluation and validation standards for the governance framework Standard What it requires in this framework Home health-specific interpretation question Construct validity Each Table 2 measure maps to a defined CoP, OASIS, or accreditation requirement Does order-authentication timeliness capture face-to-face and verbal-order compliance risk, or only documentation speed? Reliability Interrater or repeat-audit agreement on record classification Would two chart auditors classify the same referral as complete using the same field checklist?

Association with continuity outcomes An effect estimate with an uncertainty interval, adjusted for case-mix and payer How large is the association between medication reconciliation completeness and 30-day unplanned hospitalization, and how uncertain is that estimate?

Calibration Predicted versus observed escalation or readmission risk Do risk-stratification scores match observed after-hours escalation rates within strata?

Transfer across settings A held-out site, region, or payer comparison Does an authentication-timeliness protocol validated in an urban agency transfer to a rural agency with different staffing? Equity Subgroup performance by rurality, language, and payer Which patients experience the largest administrative-reliability gap, and does it close or widen after a governance change? Implementation fidelity Reach, adoption, fidelity, adaptation, cost, and sustainability Was the escalation protocol followed as designed, or adapted so heavily that fidelity cannot be assumed?

Decision value Whether the evidence changed an actual governance choice Did a measured association lead an agency to reallocate staff or revise a protocol, or was it recorded without action? These standards apply across the adoption pathway in Section 6: a component that fails construct validity or reliability at the baseline diagnostic stage should be redefined before it enters a pilot, and a component that fails transfer or equity evaluation at the protected multi-site stage should not be scaled regardless of its performance at the original pilot site.

5. Testable Propositions

The following seven propositions translate the governance framework into falsifiable claims. Each specifies the comparison, the covariates a valid test should adjust for, and the holdout condition needed to distinguish a generalizable association from a single-site artifact. None of these propositions has been tested in this manuscript; stating them in this form is intended to make future testing, and future refutation, straightforward.

Proposition 1: Agencies with a documented eligibility-verification protocol completed within twenty-four hours of referral receipt will show shorter time-to-start-of-care and a lower rate of start-of-care delay than agencies without such a protocol, after adjustment for referral source, payer type, and patient acuity at referral. A valid test should hold out at least one agency, region, or time period not used to develop the protocol and should report the association's uncertainty interval and its stability across payer type. Proposition 2: Completeness of the hospital discharge summary available at the first home health visit will be positively associated with medication reconciliation accuracy at start of care, independent of patient medication count. A valid test should preregister the reconciliation-accuracy measure before data collection and should report the association separately for verbal-order and written-order-authenticated cases, since the mechanism plausibly differs between them.

Proposition 3: Agencies with a structured after-hours escalation protocol, defined ownership, a stated response-time target, and a documented closure step, will show lower thirty-day unplanned hospitalization among their home health caseload than agencies relying on informal on-call coverage, controlling for case-mix acuity and rurality. Because rurality plausibly moderates response time independent of protocol structure, a valid test should report the association separately for rural and non-rural agencies rather than pooling them.

Proposition 4: The association between administrative reliability measures in the clinical-administrative interface domain, order-authentication timeliness in particular, and Home Health Value-Based Purchasing composite performance will be stronger among agencies with a mature quality assessment and performance improvement program than among agencies with a newly implemented one. A valid test requires an independent measure of QAPI program maturity, not inferred from the outcome itself, and should preregister the maturity threshold before examining the interaction.

Proposition 5: Structured interprofessional handoff protocols at first visit will be associated with lower documentation-latency-related delay in plan-of-care initiation than unstructured handoff practice, after adjustment for referral completeness, since incomplete referrals independently predict documentation delay and could otherwise confound the handoff association. Proposition 6: Equity monitoring stratified by rurality and preferred language will show a materially larger gap in administrative reliability measures, specifically time-to-first-visit and face-to-face documentation completeness, for rural and limited-English-proficiency patients than for other patients, and this gap will persist after adjustment for referral source. A valid test should report the gap's magnitude and uncertainty, not only its statistical significance. Proposition 7: Where a proposition in this section is not supported by data, the appropriate response is to reexamine measurement validity and mechanism specification within the relevant governance domain before concluding that the domain itself lacks value; a single null result at one site or time period should not be generalized without a holdout comparison confirming the pattern.

These seven propositions are not equally weighted for policy purposes. Propositions 1, 2, and 5 target mechanisms internal to a single episode's early days and are the most tractable to test with existing agency records, since they require only chart-level data an agency already generates for compliance purposes. Propositions 3 and 6 require after-hours and equity-stratified data that many agencies do not yet routinely structure for analysis, which is itself a finding: an agency unable to test Proposition 3 or 6 with its current records has identified a data-infrastructure gap in the escalation or equity-monitoring domain before any intervention is attempted. Proposition 4 requires an independently validated measure of QAPI program maturity that does not yet exist in a standardized form across the industry, and developing that measure is itself a contribution the field would need to make before Proposition 4 becomes directly testable; it is included here as a target for future instrument development rather than as an immediately testable claim on par with the others. Proposition 7 is procedural rather than substantive; it is included to make explicit that this paper treats a null finding as informative rather than as evidence that administrative governance itself was the wrong frame, a distinction the discussion in Section 7 returns to.

6. A Policy Adoption Pathway

Because home health agencies operate under active state licensure, federal certification, and periodic accreditation survey, a governance recommendation cannot be adopted the way a purely internal process change might be. This paper proposes a five-stage pathway suited to that regulated environment.

Governance charter. Before any measurement begins, agency leadership, clinical staff, and, where feasible, patient or caregiver representatives should agree on which of the five domains the agency will prioritize, who owns each domain's data, and what evidence will count as sufficient to change a workflow. This stage should produce a written charter, not a verbal understanding, because the accountability questions in Section 3.3 become disputable in retrospect without one. The charter should also name, explicitly, which accreditation body, CHAP, ACHC, or the Joint Commission, governs the agency's current survey cycle, since a governance change that conflicts with a standard under active survey creates avoidable compliance risk that a charter drafted without accreditation input would miss.

Baseline diagnostic. The agency measures its current performance on the Table 2 components using existing records, without yet changing any workflow, and reports missingness alongside each measure as specified in Section 4. This stage typically surfaces where a proposed governance change would address a real bottleneck versus a domain that is already functioning adequately, information a pilot cannot substitute for. A baseline diagnostic conducted mid-way through a Home Health Value-Based Purchasing performance year has the added advantage of letting an agency compare its own administrative-reliability trend against its HHVBP composite trend for the same period, without waiting for the following year's payment adjustment to reveal a problem after the fact.

Representative pilot. A limited set of components is tested in representative service lines or geographic territories, chosen to include at least one rural and one urban location where an agency serves both, since Section 2.3 and Proposition 6 give specific reason to expect the two to behave differently. The pilot should run long enough to observe a full accreditation or HHVBP performance-year cycle where feasible, and should be sized so that a null result is interpretable, an agency piloting an escalation protocol in a single five-patient rural territory will rarely generate enough after-hours events to distinguish a true null from an underpowered comparison. Protected multi-site evaluation. Where an agency operates multiple sites or participates in a multi-agency collaborative, evaluation should hold out at least one site or time period not used to develop the governance change, consistent with the propositions in Section 5, and should report uncertainty and subgroup performance rather than a single point estimate. Collaboratives that include agencies under different state licensure regimes should report results separately by state at this stage, since Section 3.4 identifies state-level Medicaid managed care variation as a boundary condition the framework does not assume away.

Scale, monitor, and revise. Expansion beyond the pilot should follow documented evidence of benefit net of implementation burden, not software availability or vendor timelines. Once scaled, the domain returns to quality and compliance governance (Section 3.1) for ongoing monitoring, and the framework should remain revisable: a governance rule that evidence contradicts should be changed, not defended. Monitoring at this stage should be scheduled to align with, rather than duplicate, the agency's existing QAPI review cadence and HHVBP reporting calendar, so that administrative-reliability monitoring becomes part of standing governance rather than a parallel reporting burden competing for the same staff time.

A timetable alone should not force advancement between these stages when measurement quality, staff capacity, or patient-safety considerations remain inadequate; the pathway is sequenced by evidence readiness, not by calendar.

6.1. Resourcing and The Burden Question

A pathway that adds measurement without acknowledging its cost is not credible in a home health setting where staff time is already constrained by productivity-based scheduling. Each Table 2 component requires a defined data source and an accountable reviewer, and for agencies without an electronic referral-management system, several of these measures, referral completeness and discharge summary completeness in particular, will initially require manual chart audit. This is a real cost, and the framework's minimum-viable-implementation principle in Section 3.1 exists specifically to bound it: an agency adopting this framework should start with the two or three components in Table 2 for which baseline diagnostic data suggest the largest gap relative to Table 1's reliability measures, rather than attempting all twelve components simultaneously. Burden should itself be tracked as an implementation-fidelity measure under Table 3, not treated as a one-time setup cost that disappears once a pilot concludes; sustained administrative measurement that consumes clinician time without a corresponding governance action, an escalation, a workflow revision, a staffing change, will not survive past the pilot stage regardless of what it shows.

7. Discussion

The review's central implication is that continuity of home health care should not be reduced to a single composite score or technology investment. Its five governance domains interact, as Section 3.2 describes, and a policy that strengthens one domain in isolation risks simply relocating the bottleneck rather than resolving it. This has a direct sequencing implication for agency leadership: definitions and workflow ownership should stabilize before an agency automates recommendations or escalation routing, because, consistent with the interoperability limits discussed in Section 3.3, poor data quality is frequently a symptom of unclear responsibility rather than a technology gap that automation alone can fix.

A further implication concerns terminology discipline. This paper deliberately avoids describing its own framework as validated, proven, or ready for national rollout, and readers who encounter that kind of language elsewhere in the administrative-reliability or care-coordination technology literature should apply the same standard this paper applies to itself: a framework paper's contribution is the specification it offers for testing, together with the conditions under which a test would count as support or refutation, not an assertion that testing has already occurred.

The review's methodological implication is that protected validation matters more in this domain than in many quality improvement contexts, because home health agencies are simultaneously subject to federal certification, state licensure, and payer-specific authorization rules that can change independently of any internal governance decision. An association observed at one agency in one payer environment may not transfer to an agency operating under a different state's Medicaid managed care rules, which is why every proposition in Section 5 specifies a holdout condition rather than relying on a single-site result. The review's equity implication follows directly from Section 2.3 and Proposition 6: administrative reliability is unlikely to be uniform across rural, limited-English-proficiency, and other structurally disadvantaged patient groups, and a governance framework that reports only agency-level averages would reproduce, rather than surface, that disparity. Equity monitoring is therefore built into Table 2 as a first-class component, not an optional add-on. A further implication concerns the relationship between this framework and the payment incentives agencies already face under Home Health Value-Based Purchasing. HHVBP rewards outcome and patient-experience measures that this paper's five domains are proposed to explain, not replace; an agency should not treat governance investment in, say, escalation response time as competing with HHVBP performance, since the propositions in Section 5 predict the two move together. Where an agency's HHVBP composite score is declining and its administrative-reliability measures show no corresponding gap, that mismatch is itself informative: it suggests either that the composite is responding to a mechanism this framework has not captured, or that the framework's measures are not yet sensitive enough to detect the underlying problem, both of which are testable rather than settled claims.

A related and frequently underestimated risk is that administrative measurement, pursued without the governance decision rights specified in Section 3.3, becomes documentation burden rather than governance. Clinicians and intake staff in home health already operate under substantial charting requirements; adding twelve tracked components without a clear answer to who reviews each one and what action follows would reproduce, at the administrative layer, the same "measurement without management" failure this paper argues has limited the existing continuity-of-care literature. The framework's insistence on named ownership, review cadence, and escalation authority for every domain in Section 3.1 is intended specifically to prevent that outcome. The review's limitations follow from its design status. No authorized dataset was analyzed, and the propositions in Section 5 are, by construction, untested. The literature synthesized in Section 2 is heterogeneous in population, setting, and outcome definition, which limits how confidently any single finding from that literature generalizes to the five-domain framework proposed here. Most of the underlying studies are also observational or single-site, which limits causal inference regardless of how the framework itself is specified; the adoption pathway's insistence on holdout comparisons in Section 6 is a direct response to this limitation rather than a generic methodological formality. Readers should treat Table 1, Table 2, and Table 3 as a measurement and governance specification to be tested, refined, and in places rejected, not as an already validated instrument.

8. Conclusion

Administrative governance is not a peripheral compliance concern for continuity of home health care; on the evidence synthesized here, it is a primary mechanism through which continuity is produced or lost, in the referral that does arrive complete or does not, the order that is authenticated on time or is not, the medication list that is reconciled before the first visit or is not, and the after-hours call that reaches someone with authority to act or does not. This paper's contribution is a five-domain governance framework, a twelve-component measurement matrix, eight evaluation and validation standards, seven falsifiable propositions, and a five-stage adoption pathway suited to a regulated, accredited industry. None of this substitutes for empirical testing. Future work should apply the propositions in Section 5 to authorized agency data, under appropriate ethics review where human-subjects protections apply, using the holdout and subgroup-reporting conditions specified throughout this paper, before any claim of effectiveness is generalized beyond the setting in which it was observed. Because the framework is deliberately scoped to governance and measurement design rather than to a single standardized intervention, it is also intended as scaffolding for more targeted work: the companion papers in this research program build directly on the domains specified here, extending the care-transition and handoff domain into a standardized, testable protocol suite (Ogundipe and Ominde, 2025) and extending the evaluation standards in Table 3 into a full psychometric quality assurance and validation framework for administrative reliability measurement (Ogundipe and Ominde, 2026). Read together, the three papers move from diagnosis, the case that administrative governance matters and how it should be organized, to intervention design, and finally to measurement validation, a sequence that mirrors how a credible policy recommendation in a regulated health care setting should actually be built.

Declarations

Ethics Approval and Consent to Participate This is a critical policy review and conceptual framework paper. No human or animal data were collected, and no patient records were accessed or analyzed for this manuscript. Any future empirical study that applies the propositions in Section 5 to agency or patient records must obtain the approvals required for its participants, records, and jurisdictions, including institutional review board approval where applicable.

Consent for Publication Not applicable.

Availability of Data and Materials No new dataset was generated or analyzed for this manuscript. The governance framework, measurement matrix, and propositions are fully reported within the manuscript itself.

Competing Interests The authors declare no competing interests.

Funding

The authors received no specific funding for this work.

Authors' Contributions Moriyanu Ogundipe and Schilder Ominde jointly conceived the governance framework, conducted the critical literature synthesis, developed the measurement matrix and propositions, and drafted, reviewed, and approved the final manuscript.