Sage Journals: Discover world-class research

Abstract

Background

Complementary and integrative health (CIH) services enhance physiological and psychological wellbeing, while potentially reducing medical costs. Despite these benefits, use of inpatient CIH services remains poorly characterized, impeding efforts to develop equitable and effective healthcare.

Objective

This retrospective case-control study examined characteristics of patients likely to receive CIH referrals and consults.

Method

Electronic health records were analyzed from patients hospitalized at a large metropolitan academic medical center from September 2022 to February 2024.

Results

Referred patients (n = 3491) were less frequently male, Asian American, non-English-speaking, and more medically complex compared to non-referred patients (n = 6982, P values <0.001). Among those referred, 72% received at least one CIH consult, with lower odds of completing a consult for male patients.

Conclusion

Disparities underscore the need for equitable CIH services access in healthcare systems. Future research will test how to broaden services to male patients, those with non-English language preference, and less medical complexity, to ensure greater benefit from holistic healthcare.

Keywords

complementary and integrative health holistic nursing mind/body pain management stress management

Background

Complementary and integrative health (CIH) combines conventional medical approaches (eg, medication) with complementary therapies (eg, massage therapy, meditation) to address people’s physical, behavioral, spiritual, and socioeconomic wellbeing holistically.¹ Over the past twenty years, leading academic medical centers and Veterans Affairs facilities have increasingly offered CIH services via a consultative model to hospitalized patients,² typically at no cost to the patient. Increased utilization in healthcare settings mirrors an upward trend in the public’s adoption of non-pharmacological strategies for health promotion and disease prevention.³ CIH research in the United States has expanded rapidly, with congressional funding for the National Center for Complementary and Integrative Health tripling between 1999 to 2024, reaching over $170 million in 2024.⁴ These trends indicate that interest for CIH is high and growing among donors, healthcare systems, and the public.

Of 5858 hospitals surveyed in 2010, 299 or 42% reported that they offer inpatient CIH services alongside conventional medical approaches.⁵ Services are typically provided to manage patient distress, anxiety, fear, and frustration that often accompany diagnosis and hospitalization⁶; and targeting mental health with behavioral interventions can reduce risk of infection and readmission.⁷ Significant gaps remain in understanding optimal delivery and implementation of CIH services, particularly in hospitals. Characterizing which patients receive services, including referral patterns and consult completion rates, is essential for identifying care gaps, developing equitable care models, enhancing patient education, and establishing efficacious treatment protocols. The following study, conducted at Tisch Hospital/Kimmel Pavilion, NYU Langone Health, a 650+ bed hospital, compared demographic and health profiles across patient groups: patients who received or did not receive a referral for CIH services, and among referred patients, those who completed a CIH intervention vs those who did not.

Method

Study Population

NYU Langone Health’s Department of Integrative Health (DIH) at Tisch/Kimmel in Manhattan has operated as a bedside consultative service since 1992. Services are ordered in the Electronic Health Record (EHR) by the attending physician, physician assistant, nurse practitioner or nurse, at no cost to patients. Thirteen DIH nurses respond to referrals by attempting a bedside consult, usually within 48 hours of receiving the order. DIH nurses are trained in multiple CIH modalities (eg, Reiki, light touch massage therapy, aromatherapy, meditation, focused breathing, guided imagery) and typically administer a combination of these to approximately 200 patients monthly. In this study, the majority (56%) of referred patients were admitted from the Emergency Department.

Study Design

This retrospective case-control study analyzed patient EHRs from hospitalizations between September 1, 2022 and February 29, 2024. To examine differences in demographic and health profiles, two main analyses were conducted comparing (1) patients who received a referral for CIH services [CIH] vs those who did not receive a referral [NR], and (2) of the referred patients, those who completed at least 1 CIH intervention vs those who did not complete any. Not all referred patients completed a CIH session due to screen outs (eg, patients with active psychosis and/or severe behavioral health symptoms, and substance use/abuse), patient unavailability (eg, asleep, with a provider, undergoing tests), or patient refusal or lack of interest.

Data Extraction

Data extraction included demographic variables (age, sex, race, preferred language) and clinical characteristics (number of comorbidities and medications, length of stay) and referral/consult outcomes. Analyses were conducted in R (v4.4.3). Duplicate patient encounters and entries lacking unique patient identifiers or admission/discharge dates were removed. Self-reported race and language variables were standardized to ensure consistency. Patients who identified as more than one race over multiple encounters were grouped as “Multiracial.” Responses to ‘preferred language’ were categorized as “English” or “Not English.”

Data Analysis

Patients were grouped according to referral status using a 1:2 frequency schema (1 referred: 2 non-referred) and matched on age using the matchit function. Since patients could have multiple hospital admissions during the study period, we analyzed demographic characteristics at the patient level (each patient included once), and clinical characteristics at the encounter level (each admission included separately to account for potential changes in patient health). EHR data on clinical characteristics were winsorized (at 0.05 and 0.95 percentiles) with the winsorize function to mitigate outlier effects. Separate multiple logistic regression models were used to compare demographic (predictors: gender, race, and preferred language) and clinical (predictors: length of stay, number of comorbidities, and number of medications) profiles between: (1) CIH and NR groups (Table 1) and (2) referred patients who completed ≥ 1 CIH consult vs those with no completed consults (Table 2). Results are reported as adjusted odds ratio (aOR) with associated 95% Confidence Interval (CI). Alpha levels were Bonferroni-corrected for multiple comparisons. Comorbidities were defined by ICD-10 codes. Medications were reported with respect to the same hospital encounter. Sample sizes were reported in the tables for each comparison.

Table 1.

Demographic and Clinical Characteristics of Referred and Non-referred Patients for Complementary and Integrative Health Services on Admission

Demographic characteristics
	Referred		Not referred		Total
	n	%	n	%	n	%	aOR (95% CI)	P
Gender
Female	2205	63.16	3972	56.89	6177	59.98	(Reference)	(Reference)
Male	1283	36.75	3008	43.08	4291	40.97	0.76 (0.70-0.82)	<0.0001^a
Non-binary	3	0.09	2	0.03	5	0.05	2.76 (0.45-16.86)	0.273
Race
White or European	2176	62.33	3869	55.41	6045	57.72	(Reference)	(Reference)
American Indian or Alaska Native	11	0.32	15	0.21	26	0.25	1.31 (0.60-2.87)	0.500
Asian or Asian American	167	4.78	574	8.22	741	7.08	0.55 (0.46-0.66)	<0.0001^a
Black or African American	551	15.78	1079	15.45	1630	15.56	0.88 (0.78-0.98)	0.024
Hispanic or Latino/a	184	5.27	359	5.14	543	5.18	1.00 (0.82-1.20)	0.964
Middle Eastern or North African	6	0.17	24	0.34	30	0.29	0.49 (0.20-1.21)	0.121
Multiracial	48	1.37	67	0.96	115	1.10	1.23 (0.84-1.79)	0.282
Native Hawaiian or Pacific Islander	4	0.11	11	0.16	15	0.14	0.66 (0.21-2.07)	0.473
Not reported	344	9.85	984	14.09	1328	12.68	0.65 (0.21-2.07)	<0.0001^a
Language
English	3314	94.98	6330	90.84	9644	92.23	(Reference)	(Reference)
Not English	175	5.02	638	9.16	813	7.77	0.59 (0.49-0.70)	<0.0001^a

Clinical characteristics
	Referred	Not referred	Total	OR (95% CI)	P
	Median (Q1-Q3)	Median (Q1-Q3)	Median (Q1-Q3)
Number of medications	43 (31-61)	7 (2-18)	19 (4-37)	1.08(1.07-1.09)	<0.0001^a
Number of comorbidities	2 (1-3)	1 (1-2)	1 (1-2)	0.97(0.90-1.05)	0.446
Length of stay (days)	7.84 (4.94-15.43)	0.20 (0.12-0.67)	0.67 (0.16-5.34)	1.79(1.71-1.87)	<0.0001^a

^aBonferroni-corrected P < 0.0001. Uncorrected P-values shown in table.

Demographic characteristics are reported at the patient-level (ie, patient only counted once, even if they have multiple hospital encounters), while clinical characteristics are reported at encounter-level (ie, patient may have multiple hospital encounters). Final sample sizes for consult data (due to screenouts and/or missing data) were as follows: Demographic (Referred [n = 3491], Not Referred [n = 6982], Total [n = 10 451]), Clinical (Referred [n = 3966], Not Referred [n = 7932], Total [n = 11 898]), Medications (Referred [n = 3966], Not Referred [n = 6912], Total [n = 10 878]), Comorbidities (Referred [n = 3652], Not Referred [n = 6810], Total [n = 10 462]), Length of stay (Referred [n = 3966], Not Referred [n = 7932], Total [n = 11 898]); Q1-Q3 = 25th percentile - 75th percentile; aOR = AdjustedOdds Ratio; CI = Confidence Interval.

Table 2.

Demographic and Clinical Characteristics of Referred Patients for Completed and Attempted Complementary and Integrative Health Consults

Demographic characteristics
	Completed		Attempted		Total		aOR (95% CI)	P
	n	%	n	%	n	%
Gender
Female	1635	64.62	414	56.95	2049	62.91	(Reference)	(Reference)
Male	891	35.22	313	43.05	1204	36.97	0.72 (0.61-0.86)	<0.0001^a
Non-binary	1	0.04	-	-	1	0.03	#	#
Unknown	3	0.12	-	-	3	0.09	#	#
Race
White or European	1497	59.17	447	61.49	1944	59.69	(Reference)	(Reference)
American Indian or Alaska Native	8	0.32	2	0.28	10	0.31	1.18(0.25-4.42)	0.833
Asian or Asian American	119	4.70	27	3.71	146	4.48	1.39 (0.90-2.17)	0.142
Black or African American	386	15.26	108	14.86	494	15.17	1.04 (0.82-1.32)	0.763
Hispanic or Latino/a	118	4.66	40	5.50	158	4.85	0.93 (0.64-1.37)	0.720
Middle Eastern or North African	5	0.20	-	-	5	0.15	#	#
Multiracial	38	1.50	3	0.41	41	1.26	3.67 (1.13-11.97)	0.031
Native Hawaiian or Pacific Islander	3	0.12	1	0.14	4	0.12	0.92 (0.09-8.87)	0.939
Not reported	356	14.07	99	13.62	455	13.97	1.10 (0.86-1.41)	0.459
Language
English	2418	95.57	679	93.40	3097	95.09	(Reference)	(Reference)
Not English	122	4.43	48	6.60	160	4.91	0.62 (0.43-0.90)	0.011

Clinical characteristics
	Completed	Attempted	Total	OR (95% CI)	P
	Median (Q1-Q3)	Median (Q1-Q3)	Median (Q1-Q3)
Number of medications	45 (33-64)	42 (27-57)	44 (33-62)	1.00 (1.00-1.01)	0.671
Number of comorbidities	2 (1-4)	2 (1-3)	2 (1-4)	0.99 (0.92-1.05)	0.669
Length of stay (days)	9.00 (5.00-17.25)	7.00 (5.00-14.00)	8.00 (5.00-16.00)	1.04 (1.01-1.06)	0.001^a

# Excluded from the logistic regression model due to separation.

^aBonferonni-corrected P < 0.01.

Final sample sizes for consult data (due to screenouts and/or missing values) were as follows: Demographic (Completed [n = 2530], Attempted [n = 727], Total [n = 3257]), Clinical (Completed [n = 2904], Attempted [n = 824], Total [n = 3728]), Medications (Completed [n = 2767], Attempted [n = 764], Total [n = 3531]), Comorbidities (Completed [n = 2582], Attempted [n = 709], Total [n = 3291]), Length of stay (Completed [n = 2904], Attempted [n = 824], Total [n = 3728]); Q1-Q3 = 25th percentile - 75th percentile; aOR = Adjusted Odds Ratio; CI = Confidence Interval.

Ethical Considerations

This study was approved by the Institutional Review Board at NYU Langone Health (study number: 23-01271).

Results

Mean age was 56 years old for referred (CIH, n = 3491) and non-referred (NR, n = 6982) groups, reflecting the age-matched study design. Multiple logistic regression identified significant associations between patient demographics and receiving a CIH referral. Male (aOR: 0.76, CI: 0.70-0.82; relative to female); Asian or Asian American (aOR: 0.55, CI: 0.46-0.66) and those who did not report their race (aOR: 0.65, CI: 0.21-2.07; relative to White or European); and non-English speaking patients (aOR: 0.59, CI: 0.49-0.70; relative to English speaking) had significantly lower odds of receiving a CIH referral (Table 1; all Bonferroni-corrected P values < 0.0001).

Relative to non-referred patients, increased medication use (aOR: 1.08, 1.07-1.09) and longer hospital stays (aOR: 1.79, CI: 1.71-1.87; Bonferroni-corrected P values < 0.0001) were also associated with significantly higher odds of receiving a referral. The median number of comorbidities was higher in referred patients (vs NR), but this difference was not statistically significant.

Among referred patients (n = 3491), 2530 (73%) received at least one CIH consult. Median time to a completed consult was 4 days (IQR: 2-8 days). The mean age for patients with completed consults was 55 years, and 57 years for those with attempted but not completed consults. Significant differences were observed only in gender between groups after correcting for multiple comparisons (Table 2). Compared to females, males had significantly lower odds of completing at least one CIH consultat (aOR: 0.72, CI: 0.61-0.86, Bonferroni-corrected P < 0.0001). Multiple logistic regression revealed that longer length of stay was associated with CIH consult completion (aOR: 1.04, CI: 1.01-1.06, Bonferroni-corrected P < 0.01), indicating that each additional day of hospitalization increased the odds of completing a CIH consult by 4%.

Discussion

This study provides a comprehensive characterization of CIH services at a large metropolitan academic medical center. Our findings reveal important demographic and clinical differences among hospitalized patients who were referred for and received CIH services. Namely, male, Asian American (and patients who did not report their race), non-English speaking patients, and those with less medical complexity were less likely to receive a referral for CIH services. While the majority (73%) of referred patients received at least 1 consult, male patients were also less likely to have completed a CIH intervention.

The gender, race, and language disparities underscore the need for more equitable access to CIH services across diverse patient populations. Similar patterns have been reported in both outpatient and inpatient settings, where the historic underrepresentation of racially/ethnically and linguistically diverse patients in CIH service access and utilization remains a recognized challenge in the field.⁸ These findings imply potential systemic barriers related to cultural, linguistic, or educational factors that might impact referrals, in addition to potential implicit bias among providers.⁹ Of note, music therapy has been successfully delivered to socioeconomically-diverse patient groups where 44% self-identified as Black and 69% reported being on Medicare/Medicaid.¹⁰ Thus, successfully integrating a wider range of therapies may aid providers in reaching more patients. Patients’ preferences undoubtedly influence their receptivity to CIH interventions, and thus, consult completion rates. A person’s familiarity with these interventions, not to mention their cultural beliefs or background, likely impacts whether they accept, refuse, or express an interest in receiving services. Though patient receptivity was not directly tested in this study, these attitudes are an important topic for future research. Addressing these gaps will require targeted provider education, improved communication, and culturally responsive care models.

Demographic characteristics among referred patients in this study mirrored that of the medical center at large, which serves a majority English-speaking patient population. However, the greater number of women who were referred and received consults cannot be explained by the medical center’s population demographics, as it serves equal numbers of men and women. Therefore, these findings reinforce the need to target men and racially and linguistically diverse groups, with potentially more robust care models that address demographic disparities in CIH access and utilization. Suggested enhancements include improving provider-patient communication, fostering peer support, and addressing biopsychosocial factors that account for social determinants of health.¹¹ Future studies also should closely examine the referral process to identify how clinical workflows can be improved to optimize equitable service delivery. As CIH services at the medical center in this study were delivered at no cost to patients, it serves as a key entry point to facilitate education on CIH for various patient groups, where otherwise in the community, patients might not have access to CIH services as readily.

With respect to clinical characteristics, more medically complex patients with significantly more medications and longer stays were more likely to be referred. Given these patients were quite ill, comorbidities, medications and length of stays were substantially high. Although patients of all medical severity types could benefit from CIH, these data suggest that services are being directed toward those with greater clinical complexity. These findings align with previous research suggesting CIH services are often utilized for chronically ill patients when conventional treatments are limited.¹² Importantly, delivering these interventions to patients with less severe presentations and early in the course of care (so-called upstream interventions)¹³ might improve their longer-term outcomes; and ideally, non-pharmacological approaches should be provided alongside established and effective conventional treatments.

Overall, this study provides a comprehensive examination of hospitalized patient characteristics throughout the referral and service delivery process from a long-established CIH program within a large metropolitan academic medical center. Limitations and barriers in the consult process have been assessed elsewhere (Millon et al, manuscript under review). This study highlighted that additional factors related to provider bias should be systematically evaluated to determine how they are impacting the ability of patients to receive these types of services. Knowledge of barriers will assist with developing targeted strategies to address disparities in access, potentially through provider education, multilingual service delivery, and culturally sensitive approaches. Even still, the fact that almost three-quarters of referred patients were visited by DIH clinicians demonstrates strong program utilization and positive reception by referring clinicians and patients. The high utilization rates suggest that nurse-led CIH services can be successfully incorporated into standard hospital care. These data provide a foundation for future research examining the effectiveness of CIH interventions for specific patient populations and clinical conditions.

Study limitations include the focus on a single institution design, which may affect generalizability. Multi-site studies and practice-based research networks like BraveNet are needed to compare patient utilization of CIH services across institutions and to examine areas of alignment and key differences in consult models.¹⁴ Future prospective studies should examine patient-reported outcomes following CIH interventions across multiple sites and explore implementation strategies to enhance equitable access to care. Nevertheless, our findings offer valuable insights for health systems seeking to implement CIH services that address the physiological and psychological needs of diverse patient populations.

Conclusion

This study characterized referral patterns and consult rates for CIH services delivered through a consultative model, providing essential insights for advancing high-quality clinical care and patient wellbeing. Female, English speaking, and medically complex patients were more likely to be referred, highlighting persistent disparities in access and utilization. Addressing these disparities is critical to ensuring equitable delivery of care. Findings lay the groundwork for improving CIH utilization and may inform the development of inpatient models at other institutions. Future research should explore how cultural, educational, and operational factors influence referral and consult patterns to optimize equitable access to CIH services.

Footnotes

ORCID iDs

Emma M. Millon

Andrea Shang

Kathleen A. DeMarco

Ethical Considerations

This study was approved by the Institutional Review Board at NYU Langone Health (study number: 23-01271).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Thank you to the Lerner family for their continued support of the Lerner Health Promotion program.

Declaration of conflicting interests

The authors declared no conflicts of interest regarding the research, authorship, and/or publication of this article.

Author Biographies

Emma M. Millon, PhD, is the Associate Director of Research in Integrative Health at NYU Langone Health and Research Assistant Professor in Child and Adolescent Psychiatry at NYU Grossman School of Medicine.

Andrea Shang, PhD, is a Senior Data Analyst in the Department of Integrative Health at NYU Langone Health.

Kathleen A. DeMarco PhD, NE-BC, CPHQ, RN, is the System Senior Director of Nursing Wellness and Resilience and System Senior Director of Lerner Health Promotion at NYU Langone Health.

References

National Center for Complementary and Integrative Health . NCCIH strategic plan FY 2021-2025: mapping a pathway to research on whole person health. 2021. https://www.nccih.nih.gov/about/nccih-strategic-plan-2021-2025

Nahin

Rhee

Stussman

. Use of complementary health approaches overall and for pain management. JAMA. 2024;331(7):613-615. doi:10.1001/jama.2023.26775

National Center for Complementary and Integrative Health . NCCIH funding: appropriations history. 2024. https://www.nccih.nih.gov/about/budget/nccih-funding-appropriations-history

Ananth

Complementary and alternative medicine survey of hospitals: Summary of results. Health Forum and Samueli Institute. 2010.

Palmer

Wehrmeyer

Florian

Raison

Idler

Mascaro

. The prevalence, grouping, and distribution of stressors and their association with anxiety among hospitalized patients. PLoS One. 2021;16(12):e0260921. doi:10.1371/journal.pone.0260921

Ford

Budworth

Lawton

Teale

O’Connor

. In-hospital stress and patient outcomes: a systematic review and meta-analysis. PLoS One. 2023;18(3):e0282789. doi:10.1371/JOURNAL.PONE.0282789

Rhee

Evans

McAlpine

Johnson

. Racial/Ethnic differences in the use of complementary and alternative medicine in US adults with moderate mental distress: results from the 2012 national health interview survey. J Prim Care Community Health. 2016;8(2):43-54. doi:10.1177/2150131916671229

Holden

McGregor

Thandi

, et al. Toward culturally centered integrative care for addressing mental health disparities among ethnic minorities. Psychol Serv. 2014;11(4):357-368. doi:10.1037/a0038122

Hall

Chapman

Lee

, et al. Implicit racial/ethnic bias among health care professionals and its influence on health care outcomes: a systematic review. Am J Publ Health. 2015;105(12):e60-76. doi:10.2105/AJPH.2015.302903

10.

Rodgers-Melnick

Rivard

Block

Dusek

. Effectiveness of medical music therapy practice: integrative research using the electronic health record: rationale, design, and population characteristics. J Integr Complement Med. 2024;30:57-65. doi:10.1089/jicm.2022.0701

11.

Chao

Thompson-Lastad

Swedlow

, et al. Group-based integrative pain management in primary care: a study protocol for multilevel interventions to address health disparities. Glob Adv Integr Med Health. 2024;13:27536130241302043. doi:10.1177/27536130241302043

12.

Armstrong

Lanni

Anderson

Patricolo

. Integrative medicine and the oncology patient: options and benefits. Support Care Cancer. 2018;26(7):2267-2273. doi:10.1007/s00520-017-4007-y

13.

National Academies of Sciences, Engineering, and Medicine . Achieving whole health: a new approach for veterans and the nation. 2023. https://nap.nationalacademies.org/catalog/26854/achieving-whole-health-a-new-approach-for-veterans-and-the

14.

Dusek

Gao

Kim

, et al. Patients receiving integrative medicine effectiveness registry (PRIMIER) of the BraveNet practice-based research network: outcomes of the PRIMIER cohort. Compl Ther Med. 2022;71:102904. doi:10.1016/j.ctim.2022.102904

Characteristics of Hospitalized Patients Referred for Complementary and Integrative Health Services at a Large Metropolitan Academic Medical Center

Abstract

Background

Objective

Method

Results

Conclusion

Keywords

Background

Method

Study Population

Study Design

Data Extraction

Data Analysis

Ethical Considerations

Results

Discussion

Conclusion

Footnotes

ORCID iDs

Ethical Considerations

Funding

Declaration of conflicting interests

Author Biographies

References