Factors Associated with Timing of the Start-of-care Nursing Visits in Home Health Care

Abstract

Objectives

Home health care patients have critical needs requiring timely care following hospital discharge. Although Medicare requires timely start-of-care nursing visits, a significant portion of home health care patients wait longer than two days for the first visit. No previous studies investigated the pattern of start-of-care visits or factors associated with their timing. This study’s purpose was to examine variation in timing of start-of-care visits and characterize patients with visits later than two days post-discharge.

Design

Retrospective cohort study

Setting/participants

Patients admitted to a large, Northeastern United States, urban home health care organization during 2019. The study included 48,497 home care episodes for 45,390 individual patients.

Measurement

We calculated time to start-of-care from hospital discharge for two patient groups: those seen within two days versus those seen > 2 days post-discharge. We examined patient factors, hospital discharge factors, and timing of start-of-care using multivariate logistic regression.

Results

Of 48,497 episodes, 16,251 (33.5%) had a start-of-care nursing visit > 2 days after discharge. Increased odds of this timeframe were associated with being Black or Hispanic and having solely Medicaid insurance. Odds were highest for patients discharged on Fridays, Saturdays, and Mondays. Factors associated with visits within 2 days included surgical wound presence, urinary catheter, pain, 5 or more medications, and intravenous or infusion therapies at home.

Conclusion and implications

Findings provide the first publication of clinical and demographic characteristics associated with home health care start-of-care timing and its variation. Further examination is needed and adjustments to staff scheduling and improved information transfer are two suggested interventions to decrease variation.

Introduction

Home health care services play an important role in the U.S. health care system. Since the early 1970s, the health care delivery system has changed dramatically with an increasing number of clinically complex patients treated in post-acute care settings, including home health care.^1,2 Home health care services consist of a series of home visits conducted by registered nurses and other healthcare providers (e.g., physical therapists, social workers, occupational and speech therapists). As the population continues to age, the demand for home health care increases.³ According to the Centers for Medicare & Medicaid Services (CMS), home health care expenditure was over $102 billion in 2018, indicating a 30% increase in spending compared to 5 years ago.⁴ Studies indicate home health care services improve patient outcomes and reduce costs.^5–7 Utilization of home health care is expected to continue to expand in the near future.⁸

This study focuses on patients transitioned to home health care from hospitals, who constitute up to 70% of the home health care population.⁹ They tend to have treatment needs related to pain management, medication management, wound care, and other sub-acute/chronic medical management that require timely care soon after hospital discharge.^10,11 Recent nationwide studies show that about one-in-five home health care patients are rehospitalized during home health care services.⁹ Up to 70% of these rehospitalizations happen during the first two weeks of the home health care episode making timely visits important.^10,12–15

At start-of-care the home health care clinician assesses the patient’s clinical condition, reconciles medications, evaluates the extent of caregiver support, reviews hospital discharge instructions, and designs a personalized care plan for the patient’s recovery.¹⁶ Timing of the start-of-care visit has important patient care implications, especially for patients with unstable or urgent health needs. A recent study demonstrated that prioritizing clinically complex patients for the start-of-care visit can reduce rehospitalizations by 50%.¹⁵

Other studies have also demonstrated that the timing of home health care services, such as frontloading of visits after admission, are associated with significantly lower rehospitalization rates.^10,14,17,18 The Center for Medicare and Medicaid Services (CMS) encourages timely visits through a requirement (described in the CMS Conditions of Participation) which states that “the initial assessment visit must be held either within 48 hours of referral, or within 48 hours of the patient’s return home, or on the physician-ordered start-of-care date”.¹⁹

Given the volume of patients transitioning to home health care, it is difficult to meet the CMS timely start-of-care standard for all patients. Best practice would be to prioritize those with the greatest needs. However, without a picture of the patients’ needs or tools to identify risk, patients who may benefit from timely attention may not receive priority. Our extensive literature search identified no previous studies that investigated the pattern of start-of-care visits or what factors are associated with their timing. In an effort to understand this important quality event, this study’s purpose was to examine the variation in timing of the start-of-care visits and to characterize patients with start-of-care occurring greater than two days after hospital discharge.

Methods

Study dataset

The study dataset included all patients admitted to home health care services by a health care provider (e.g., registered nurse, physical therapists) at a large urban home health care organization between 01/01/2019 and 12/31/2019. Data were extracted from the Outcome and Assessment Information Set (OASIS, Version D). OASIS is a federally required and standardized assessment tool for adult home health care patients. Nurses use OASIS to comprehensively assess nearly 100 items including demographics, clinical status (e.g., medical history and comorbidities), home environment, informal caregivers, functional status, and health service utilization.²⁰ The study was approved by the Institutional Review Boards of the participating institutions.

Study population

This study focused on the timing of the start-of-care nursing visits in all home health care episodes for patients transitioning from acute care. We focused on patients coming from the hospital as they enter home care with more acute needs and greater risk of re-hospitalization.²¹ A patient could have more than one home health care episode; there were 48,497 home health care episodes for 45,390 individual patients. All analyses were conducted at the home health care episode level.

Definition of start-of-care nursing visit timing

Based on the CMS definition for a timely initial assessment,²² we chose the most conservative/clinically important measure as the difference in days between the hospital discharge date (OASIS item “M1005: Inpatient Discharge Date”) and the OASIS completion date (OASIS item “M0090: Date Assessment Completed”). Timing of the visit was categorized as patients who received the start-of-care visit ≤ 2 days or > 2 days after hospital discharge.

Potential patient-related factors associated with timing of the start of home health care services

The admission OASIS assessment with additional diagnosis information from the plan of care provided the patient-related factors (Appendix 1). Patient-related factors included: 1) demographics: age, gender, and race; 2) socioeconomic status: type of healthcare insurance and living conditions – e.g., patient lives alone; 3) current medical conditions: comorbidities and therapies at home; 4) hospitalization risk factors: previous 12 month fall history and currently taking 5 or more medications; 5) sensory status: vision limitations and frequency of pain; 6) integumentary status: pressure ulcers and other wounds; 7) respiratory status; 8) elimination status: history of urinary tract infection, and presence of urinary catheter or bowel ostomy; 9) cognitive function; 10) dependency for management of oral medications; and 11) functional disability.

Common comorbid conditions were congestive heart failure, cardiac arrhythmias, hypertension, chronic pulmonary disease, diabetes, renal failure and cancer. The Elixhauser Comorbidity Score: a composite score of over 30 individual comorbidities associated with mortality based on the individual ICD-10-CM (International Classification of Diseases, Tenth Revision, Clinical Modification) diagnosis codes provided a measure of disease burden.^23–26

Functional disability refers to the patient’s inability to conduct activities independently documented in OASIS. The assessment consists of activities of daily living (ADL)/instrumental activities of daily living (IADL) function. ADL items include grooming, dressing upper, dressing lower, bathing, toileting, transferring, ambulating, and eating. IADL item included meal preparation. “ADLs Needed” was defined as the sum of the binary ADL/IADL items (total ranged from 0–9). When the response for ADL/IADL item is 0 (indicating no issues) the binary indicator was 0, otherwise it was 1 (indicating moderate or significant issues). “ADLs Severity” was defined as the sum of the response categories of the level of dependency in ADL/IADL items (total ranged from 0–38). An independent item was 0 while the most severe level of dependency ranged from 3 to 6 (e.g., 3 for grooming, 4 for toilet transferring, 5 for transferring, or 6 for ambulating).²⁷

Potential hospital discharge factors associated with timing of the start of home health care services.

A single hospital discharge factor was the day of the week on which patients were discharged from the hospital.

Statistical analysis

Descriptive statistics were used to calculate the proportion of patients with start-of-care ≤ 2 days or > 2 days from hospital discharge. Characteristics of patients in the two groups were compared using student t-test or chi-squared tests, when appropriate.

Bivariate logistic regression analysis identified the patient characteristics or discharge day of the week associated with visits > 2 days. Factors that differed at a 5% level were incorporated into a multivariate logistic regression model to examine the relationship between associated factors and start-of-care occurring > 2 days, adjusted for other covariates including the hospital discharge factor. Due to multicollinearity (i.e., high degree of correlation) between Elixhauser Comorbidity Score and seven individual comorbidities, only Elixhauser Comorbidity Score was included in the multivariate analysis. For all analyses, factors with p-values less than 0.05 were considered statistically significant. Odds ratios (OR) were calculated with 95% confidence intervals (CI). All analyses were done using R (Foundation of Statistical Computing, Vienna).

Results

Timing of start-of-care visits

One-third of home health care episodes (33.5%, 16,251) received their start-of-care visits later than day 2. Figure 1 shows the proportion of days between hospital discharge to start-of-care home health care nursing visit. On average, there were 2.61 days (standard deviation [SD] =2.62) between hospital discharge to start-of-care home health care nursing visit.

Figure 1.

Days between hospital discharge to homecare first nursing visit

Characteristics of the cohort

Characteristics of patients with start-of-care ≤ 2 days versus those with > 2 days home health care visits are summarized in Table 1.

Table 1.

Comparison of Factors between Patients who Received the Start-of-Care Visit ≤ 2 days and > 2 days after Hospital Discharge

	Start-of-Care Visit ≤ 2 days (n = 32,246)	Start-of-Care Visit > 2 days (n = 16,251)	Differences	P-value*
1. Patient related factors
1.1. Demographics
Age (%)				<0.0001
Age < 65	40.1	34.4	+5.7
Age ≥ 65	59.9	65.6	−5.7
Sex (%)
Female	57.5	58.9	−1.4	0.003
Race (%)				<0.0001
Non-Hispanic White	49.4	31.6	17.8
Non-Hispanic Black	20.8	30.0	−9.2
Hispanic	20.8	28.9	−8.1
Other	9.0	9.5	−0.5
1.2. Socio-economic Status
Type of insurance (%)				<0.0001
Dual eligibility	9.7	16.3	−6.6
Medicare FFS only	28.4	24.6	+3.8
Medicaid FFS only	0.7	1.2	−0.5
Any managed care	29.4	39.8	−10.4
Other (e.g., private)	31.8	18.1	+13.7
Living Condition (%)				<0.0001
Living with others (Congregate or With Other)	75.1	73.2	+1.9
Living alone	24.9	26.8	−1.9
1.3. Current Condition
Elixhauser Comorbidity Score [mean, (SD)]	1.75 (1.29)	2.21 (1.33)	−0.46	<0.0001
Active Diagnoses (%) **
Congestive heart failure	11.1	19.2	−8.1	<0.0001
Cardiac arrhythmias	9.1	12.0	−2.9	<0.0001
Hypertension	55.9	66.2	−10.3	<0.0001
Chronic pulmonary disease	12.5	16.3	−3.8	<0.0001
Diabetes	27.6	38.8	−11.2	<0.0001
Renal failure	8.8	14.9	−6.1	<0.0001
Cancer	16.4	16.4	0	<0.0001
Therapies the patient receives at home (either intravenous/infusion therapy or enteral nutrition) (%)	4.0	3.0	1.0	<0.0001
1.4. Risk for Hospitalization (%)
History of falls in the past 12 months	11.4	15.8	−4.4	<0.0001
Unintentional weight loss in the past 12 months	5.1	6.4	−1.3	<0.0001
Multiple hospitalizations (2 or more) in the past 6 months	20.4	29.3	−8.9	<0.0001
Multiple emergency department visits in the past 6 months	12.8	20.4	−7.6	<0.0001
Decline in mental, emotional, or behavioral status	7.4	9.5	−2.1	<0.0001
Reported or observed history of difficulty complying with any medical instructions	9.9	11.8	−1.9	<0.0001
Currently taking 5 or more medications	80.0	81.1	−1.1	<0.0001
1.5. Sensory Status
Vision (%)				<0.0001
Normal vision	83.4	78.6	+4.8
Impaired vision (partially or severely)	16.6	21.4	−4.8
Frequency of Pain (%)				<0.0001
Patient has no pain	12.7	19.9	−7.2
Patient has pain, but not all of the time	77.7	74.1	+3.6
All of the time	9.6	6.0	+3.6
1.6. Integumentary (%)
Having at least one Unhealed Pressure Ulcer at Stage II or Higher	2.1	2.8	−0.7	<0.0001
Having Stasis Wound	0.96	1.33	−0.37	<0.0001
Having Surgical Wound	60.9	33.4	+27.5	<0.0001
1.7. Respiratory Status (%)
Short of Breath				<0.0001
Never	52.6	44.2	+8.4
When walking more than 20 feet/climbing stairs	29.4	33.8	−4.4
With moderate/minimal exertion or at rest	18.0	22.0	−4.0
1.8. Elimination
Urinary Tract Infection in the past 14 days (%)				<0.0001
No/Unknown	93.9	91.5	+2.4
Yes or Patient on prophylactic treatment	6.1	8.5	−2.4
Urinary Incontinence or Urinary Catheter Presence (%)				<0.0001
None	72.8	62.6	+10.2
Incontinent	23.2	34.0	−10.8
Required Urinary Catheter	4.0	3.4	+0.6
Ostomy for Bowel Elimination (%)				<0.0001
None	98.0	98.5	−0.5
Having Ostomy was not related to an inpatient stay	2.0	1.5	+0.5
1.9. Neuro, Emotional, and Behavioral Status
Cognitive Functioning (%)				<0.0001
Alert/oriented or Prompting	95.3	92.4	+2.9
Requires assistance or totally dependent	4.7	7.6	−2.9
When Confused (%)				<0.0001
Never	70.7	62.7	+8.0
In new or complex situations	25.3	30.4	−5.1
On awakening and/or at night only, or consistently	4.1	6.9	−2.8
1.10. Medications
Management of Oral Medications (%)				<0.0001
Independent or No oral medications	30.6	28.0	+2.6
Preparation or reminders needed	34.2	35.6	−1.4
Dependent	35.2	36.5	−1.3
1.11. ADLs / IADLs
ADL Needed [mean, (SD)]	1.81	2.0	−0.19	<0.0001
ADL Severity [mean, (SD)]	5.69	6.43	−0.74	0.58
2. Hospital discharge related factors
2.1. Discharge day of week (%)				<0.0001
Sunday	8.3	6.0	+2.3
Monday	11.8	13.7	−1.9
Tuesday	14.3	15.2	−0.9
Wednesday	16.2	16.1	+0.1
Thursday	17.6	16.4	+1.2
Friday	19.9	21.4	−1.5
Saturday	11.9	11.2	+0.7

Note:

^*student-t test for continuous variables and chi-squared tests for categorical variables as applicable.

^**The presented individual active medical conditions were not included in the multivariate analysis due to multicollinearity with the Elixhauser Comorbidity Score.

On average, patients with patients with > 2 days home health care nursing visits were older, more likely to be female, and more likely to be non-Hispanic Black or Hispanic. Approximately 40% of patients hold Medicare fee for service, Medicaid fee-for-service or dual eligibility. Two-thirds of all patients were living with others, but patients with start-of-care > 2 days were more likely to live alone. Elixhauser comorbidity score representing the disease burden was higher in the patients with start-of-care > 2 days (1.75 vs. 2.21, p <0.0001). Patients with start-of-care > 2 days were more likely to have hospitalization risk factors, including multiple hospitalizations/ emergency department visits in the past 6 months. In addition, shortness of breath, history of urinary tract infection, or urinary incontinence were more frequent in patients with delayed start-of-care. Patients with start-of-care > 2 days had decreased cognitive function and a status of confusion, and were more dependent with oral medication management than patients with start-of-care ≤ 2 days home health care visits.

Patients with start-of-care ≤ 2 days were more likely to hold Medicare fee-for-service or other types of insurance, such as private insurance (28.4% vs. 24.6% and 31.8% vs. 18.1%, respectively, p <0.05). Patients with start-of-care ≤ 2 days had higher frequency of receiving intravenous/infusion therapy or enteral nutrition at home, reported more pain, and had more surgical wounds. Patients with start-of-care ≤ 2 days were more likely to have an ostomy or urinary catheter. The average number of ADL/IADL dependencies was higher in the patients with start-of-care ≤ 2 days (8.02 vs. 7.91; p <0.001). However, ADLs severity was not significantly different between the patients with start-of-care ≤ 2 days and > 2 days (14.62 vs 14.59, p = 0.58).

Hospital discharge factors

Over 20% of patients were discharged from acute care on a Friday; while the fewest discharges occurred on Sundays (7.5%) as shown in Figure 2. More frequent start-of-care > 2 days visits occurred when patients were discharged from hospitals on Mondays, Tuesdays, and Fridays.

Figure 2.

The proportion of hospital discharges between start-of-care ≤ 2 days group and a start-of-care > 2 days group by day of week

Factors associated with timing of start-of-care visits

In the bivariate analyses, all factors (except ADL severity) were significantly different between the patients with start-of-care ≤ 2 days and those with > 2 days. As a result, all factors were included in the multivariable logistic regression.

Only factors significantly associated with risk for start-of-care > 2 days after adjusted regression analysis are shown in Table 2. Patients with surgical wounds were 56% less likely to have start-of-care > 2 days compared with patients without surgical wounds (OR, 0.44 [95% CI, 0.42–0.46]). Patients who reported having the highest frequency of pain (i.e., “all of the time”) had the second-lowest odds for start-of-care > 2 days compared to patients without pain (OR, 0.69 [95% CI, 0.63–0.76]). Being 65 years of age or older was associated with a slight decrease in odds of start-of-care > 2 days compared to those under 65 (OR, 0.98 [95% CI, 0.9–0.99]).

Table 2.

The associations between a timing of start-of-care and patient-related/hospital discharge-related factors^†

	Multivariable analysis
	Adjusted OR [95% CI]
1. Patient related factors
1.1. Demographics
Age
Age < 65	Reference
Age ≥ 65	0.98 [0.9–0.99] *
Race
Non-Hispanic White	Reference
Non-Hispanic Black	1.76 [1.67–1.86] **
Hispanic	1.61 [1.53–1.7] **
Other	1.4 [1.3–1.51] **
1.2. Socio-economic Status
Type of insurance
Dual eligibility	1.29 [1.21–1.39] **
Medicare FFS (fee for service) only	Reference
Medicaid FFS only	1.56 [1.26–1.92] **
Any managed care	1.26 [1.2–1.33] **
Other (e.g., private)	0.86 [0.81–0.92] **
1.3. Current Condition
Therapies the patient receives at home (either intravenous/infusion therapy or enteral nutrition)	0.71 [0.64–0.8] **
1.4. Risk for Hospitalization
History of falls in the past 12 months	1.21 [1.14–1.28] **
Multiple hospitalizations (2 or more) in the past 6 months	1.14 [1.08–1.21] **
Multiple emergency department visits in the past 6 months	1.17 [1.09–1.24] **
Currently taking 5 or more medications	0.91 [0.87–0.96] *
1.5. Sensory Status
Patient has no pain	Reference
Patient has pain, but not all of the time	0.86 [0.81–0.91] **
All of the time	0.69 [0.63–0.76] **
1.6. Integumentary
Having Surgical Wound	0.44 [0.42–0.46] **
1.7. Respiratory Status
Short of Breath
Never	Reference
When walking more than 20 feet/climbing stairs	1.11 [1.05–1.16] **
With moderate/minimal exertion or at rest	1.02 [0.97–1.08]
1.8. Elimination
Urinary Tract Infection in the past 14 days
No/Unknown	Reference
Yes or Patient on prophylactic treatment	1.1 [1.02–1.19] *
Urinary Incontinence or Urinary Catheter Presence
None	Reference
Incontinent	1.13 [1.08–1.19] **
Required Urinary Catheter	0.78 [0.7–0.87] **
1.10. Medications
Management of Oral Medications
Independent or No oral medications	Reference
Preparation or reminders needed	1.03 [0.98–1.09]
Dependent	0.93 [0.88–0.99] *
1.11. ADL/IADL
ADL Needed [mean, (SD)]	0.94 [0.93–0.95] **
2. Hospital discharge related factors
2.1. Discharge day of week
Sunday	Reference
Monday	1.29 [1.18–1.42] **
Tuesday	1.18 [1.07–1.3] *
Wednesday	1.16 [1.06–1.27] *
Thursday	1.21 [1.02–1.23] *
Friday	1.32 [1.21–1.44] **
Saturday	1.33 [1.2–1.46] **

Notes:

^†. The multivariate adjusts were conducted for all variables with statistical significance from the bivariate analysis. However, only estimates of variables with statistical significance from multivariate logistic regression are presented in Table 2;

^*= p-value <0.05;

^**= p-value <0.0001

Non-Hispanic Black (OR, 1.76 [95% CI, 1.67–1.86]) and Hispanic (OR, 1.61 [95% CI, 1.53–1.7]) patients had the highest odds for start-of-care > 2 days compared to non-Hispanic White patients. Odds of start-of-care > 2 days also varied by the type of health insurance coverage. Dually eligible patients, Medicaid fee-for-service recipients and any managed care patients had increased risk of start-of-care > 2 days compared to recipients of Medicare fee-for-service (Dual eligible – OR, 1.29 [95% CI, 1.21–1.39]; Medicaid fee-for-service – OR, 1.56 [95% CI, 1.26–1.92]; managed care – OR, 1.26 [95% CI, 1.2–1.33], respectively). Recipients of other types of insurance, including private insurances, were less likely to have start-of-care > 2 days (OR, 0.86 [95% CI, 0.81–0.92]).

Patients with more ADL limitations and who needed higher levels of assistance had lower risk for the start-of-care > 2 days (OR, 0.94 [95% CI, 0.93–0.95]). Patients discharged from hospitals on Sundays had lowest odds for start-of-care > 2 days compared to those discharged on all the other days of the week, especially on Fridays and Saturdays.

Patients identified at a higher risk for hospitalization due to history of falls in the past 12 months, and/or multiple hospitalizations and emergency department visits in the past 6 months had higher odds of start-of-care > 2 days compared to patients without these risk factors (OR, 1.21 [1.14–1.28], OR, 1.14 [1.08–1.21], and OR, 1.17 [1.09–1.24], respectively).

Discussion

This study is the first to investigate the extent of clinical and demographic characteristics associated with the timing of start-of-care in home health care and the extent of variation. We found that about one-in-three patients in the large patient sample had a start of nursing care visit > 2 days. This number is higher than CMS’s publicly available reporting, which indicates only 5% of patients nationwide experience start-of-care later than 48-hours after returning to the home.²⁸ The differences between our study findings and CMS data might be attributed to the different populations and definitions of visit timeliness. CMS reporting covers all patients admitted to home health care, including patients referred from community and other post-acute care settings. In this study, only patients discharged from hospitals were included. More importantly, our start-of-care visit timeliness calculation was based on the most conservative CMS definition of hospital discharge date rather than the home health care referral date. Hence, the study’s calculation reflects the number of days patients spent in the community after discharge, rather than counting from the home health care referral dates which could be received any time before or after hospital discharge. Further, some agencies may clock the referral date from when the case is accepted into the agency, giving them more time than the more conservative calculation of time since discharge. Our results suggest that different definitions of start-of-care visit timeliness may influence which factors are associated with the outcome. Therefore, further study is warranted on the comparison of the timeliness of start-of-care as calculated from the discharge date versus referral date to examine the differences in outcomes.

The study identified several patient level factors that decreased the likelihood of start of nursing care visit > 2 days. These factors included presence of surgical wounds, having pain, receiving therapies at home, presence of urinary catheter, taking 5 or more medications, dependency in managing oral medications, higher ADLs needed, and older age. A possible explanation of these factors is that perhaps home health care organizations may have the information regarding the above mentioned factors before they arrange start-of-care visits.²⁹ Or these factors which decreased the risk of start-of-care > 2 days might be co-related to each other. For example, patients having surgical wounds can complain of more pain, or older aged patients are more dependent on the management of oral medication or require more ADLs assistance. In addition, wounds and multiple medications were also identified in recent studies, including the study that created the clinical decision support tool “PREVENT” which prioritizes patients for the start-of-care visit.³⁰ Our findings regarding wounds and multiple medications are also supported by the hospitalization risk factors in home health care literature.³¹

Several factors were associated with the risk of start-of-care > 2 days. Patient race was a significant factor associated with higher odds of start-of-care > 2 days; Non-Hispanic Black and Hispanic patients had greater than 60% higher odds of start-of-care > 2 days. This finding is similar to a recent nationwide analysis of Medicare data that found race disparities in home health care (i.e., Black or Hispanic patients received home health at lower rates than White patients).³² This same study found that disadvantaged patients living in high-unemployment zip codes were more likely to have start of nursing care visit > 2 days.³² Further investigation of staff allocation by geographical area is warranted to better understand the association between geographic area (e.g., urban/rural, or borough), race, and start-of-care visit timing. We also identified that type of insurance was a significant factor associated with odds of start-of-care > 2 days. It is possible that start of nursing care visit > 2 days for those with managed care insurance coverage may be due to the time it takes to secure an authorization for services.

Unexpectedly, we found that patients identified as being at higher risk for hospitalization (due to falls history, and recent multiple hospitalizations and emergency department visits) were more likely to have start-of-care > 2 days compared to patients without these risk factors. In addition, several other clinical factors, including incontinence or shortness of breath, were also associated with start of nursing care visit > 2 days. While a home health care referral may have some high-level clinical information that will inform the plan of care, such as the need for wound care, therapies, medication management, detailed clinical information is often not available prior to admission. A recent study found that prior to the start-of-care visit nurses tend not to have needed patient clinical information,³³ including patient risk for hospitalization. It is not until the first visit when the healthcare providers (e.g., nurse or physical therapist) collect the OASIS that these items become known. The lack of detailed of upfront clinical information limits informed decision-making about how to prioritize visit schedules.

Patients discharged on Sundays had the lowest odds of start-of-care > 2 days nursing visits compared to other days of the week. Patients discharged on Mondays, Fridays and Saturdays had the highest odds of start-of-care > 2 days nursing visits. This finding may be explained by variations in workload or staffing ratios during weekends and weekdays. During weekends, the staffing levels are generally low while home health care admissions continue. During early weekdays, the organization may experience highest volume of referred patients. Therefore, it is challenging to meet the demand for timely visits within 2 days for different reasons. To address this issue, the home health care organization may need to adjust their staffing approaches.

Timing of the start-of-care has become an increasing problem due to the increase in the number of patients being referred to home health care. Referral systems are overloaded along with hospital discharge staff, home health in-take staff, and home health clinical staff. Start-of-care may take longer in the rural areas than the urban areas due to limited home care clinical staff and the increased drive time to and between patients requiring more time.³⁴ Initial OASIS clinical assessment visits, when done by qualified staff, can take longer than the usually allocated 1 hour of visit time depending on the complexity and care needs of the patient.³⁵ These issues are all factors that could delay the initial start-of-care of patients. Therefore, the home health care organizations may want to examine their start-of-care visit patterns and discuss improving their processes aimed at improving the vist timeliness and subsequent patient outcomes.

Limitations

This study is limited to 2019 data from one large, urban home health care agency in the Northeastern United States. Results might not be generalizable to other home health care locations. All associations identified are correlations and do not identify causality.

Further research

Further analysis is needed to understand the associations between visit timing and race, potentially involving detailed geographic and census information with home health care agency staffing ratios; and the potential interrelated factor of late referral. In addition, home health care agencies’ reasons for start of nursing care visit > 2 days should be investigated, including agencies’ ability to contact patients/caregivers to schedule home care start-of-care visits, or patients with higher ADL status not staying at home for scheduled start-of-care visits. Future qualitative work may identify the role patient preference plays in scheduling care as well.

Conclusions and implications

This study was the first study to examine patient characteristics associated with timing of start-of-care nursing visits in home health care. One-third of patients admitted to a large urban home health care organization had start-of-care visits on later than day 2. Several clinical and demographic characteristics were associated with this timing and others have shown this is associated with disparate clinical outcomes. Further investigation is needed to determine why there is variation in timing, which factors are modifiable, and the associated consequences of timing during the transition from hospital to home care.

Impact Statement.

We certify that this work is the first to examine patient characteristics associated with timing of the start-of-care nursing visits in home health care. Certain clinical and demographic characteristics were associated to timing of the start-of-care and have the possibility of leading to disparate clinical outcomes.

Abbreviations

CMS

Centers for Medicare & Medicaid Services

PREVENT

Priority for the First Nursing Visit Tool

OASIS

Outcome and Assessment Information Set

ADL/IADL

Activities of Daily Living/Instrumental Activities of Daily Living function

OR

Odds Ratios

CI

Confidence Intervals

SD

Standard Deviation

Appendix 1:

OASIS and administrative items reflecting patients’ characteristics

Category	Variable	Data Source	Type of Data
Demographics	Age	Patient administrative Data	Binary:  Age < 65  Age ≥ 65
	Sex	OASIS M0069	Binary: Female/Male
	Race/ethnicity	M1040	Categorical:
	American Indian or Alaska Native		Non-Hispanic
	Asian		White
	Black or African-American		Non-Hispanic
	Hispanic or Latino		Black
	Native Hawaiian or Pacific Islander		Hispanic
	White		Other
Socio-economic Status	Payer	M0150	Categorical:
	Medicare FFS		Dual eligibility
	Medicaid FFS		Medicare FFS only
	Medicare Managed Care		Medicaid FFS only
	Medicaid Managed Care		Any Managed Care
	Private Insurance		Other (e.g., private)
	Others
	Living Condition	M1100	Binary:
	Patient lives alone		Living alone
	Patient lives with other person(s) in the home		Living with others
	Patient lives in congregate situation (e.g., assisted living, residential care home)
Current Condition	Elixhauser Comorbidity Score	Electronic health records (ICD-9-CM or ICD-10-CM codes)	Continuous
	Diagnoses	M1020, M1022, M1024. Plan of care	Binary: Yes/No
	Congestive heart failure
	Cardiac arrhythmias
	Hypertension
	Chronic pulmonary disease
	Diabetes
	Renal failure
	Cancer
	Therapies	M1030	Binary: Yes/No
	Intravenous or infusion therapy
	Parenteral nutrition
	Enteral nutrition
Risk for Hospitalization	History of falls in the past 12 months	M1032	Binary: Yes/No
	Unintentional weight loss in the past 12 months
	Multiple hospitalizations (2 or more) in the past 6 months
	Multiple emergency department visits in the past 6 months
	Decline in mental, emotional, or behavioral status
	Reported or observed history of difficulty complying with any medical instructions
	Currently taking 5 or more medications
Sensory status	Vision	M1200	Binary:
	Normal vision		Normal impaired vision
	Partially impaired: cannot see medication labels or newsprint, but can see obstacles in path, and the surrounding layout; can count fingers at arm’s length.
	Severely impaired: cannot locate objects without hearing or touching them, or patient nonresponsive.
	Frequency of Pain	M1242	Categorical:
	Patient has no pain		Patient has no pain
	Patient has pain that does not interfere with activity or movement		Patient has pain, but not all of the time
	Less often than daily		All of the time
	Daily, but not constantly
	All of the time
Integumentary Status	Having at least one Unhealed Pressure Ulcer at Stage II or Higher	M1306	Binary: Yes/No
	Having a Stasis Ulcer?	M1330	Binary: Yes/No
	Having Surgical Wound	M1340	Binary: Yes/No
Respiratory Status	Short of Breath	M1400	Categorical:
	Never		Never  When walking more than 20 feet/climbing stairs
	When walking more than 20 feet/climbing stairs
	With moderate exertion
	With minimal exertion
			With
			moderate/minimal exertion or at rest
	At rest
Elimination Status	UTI Treatment in Past 14 Days	M1600	Binary: Yes/No
	Urinary Incontinence or Urinary Catheter	M1610	Categorical:
	Presence		None
	None
			Incontinent
	Incontinent
			Required Urinary Catheter
	Required Urinary Catheter
	Ostomy for Bowel Elimination	M1630	Binary:
	None		None
	Ostomy was not related to an inpatient stay		Ostomy was not related to an inpatient stay
	Ostomy was related to an inpatient stay
Neuro/Emotion	Cognitive Functioning	M1700	Binary:
al/Behavioral Status	Alert/oriented		Alert/oriented or Prompting
	Requires prompting
	Requires assistance
			Requires assistance or totally dependent
	Requires considerable assistance
	Totally dependent
	When Confused	M1710	Categorical:
	Never		Never
	In new or complex situations/On awakening or at night only		In new or complex situations
	During the day and evening but not constantly
			On awakening and/or at night only, or consistently
	Constantly
	Patient nonresponsive
Medication & Medication Management	Management of Oral Medications	M2020	Categorical:  Independent or no Oral Meds  Preparation needed or reminders needed  Dependent
	Independent
	Preparation needed
	Reminders needed
	Dependent
	No Oral Meds
ADLs / IADLs	ADL Needed  The sum of the binary ADL/IADL items: when the response for ADL/IADL item is 0 (indicating no issues) then the binary indicator was 0, otherwise it was 1 (indicating moderate or significant issues).	M1800~M1890	Continuous
	ADL Severity  The sum of the response categories of the level of dependency in ADL/IADL items: totally independent item was 0 while a totally dependent item varied from 3 to 6.	M1800~M1890	Continuous