Patient-sharing Network Research Articles

Oncology Physician Turnover in the United States Based on Medicare Claims Data

Objective: Physician turnover rates are rising in the United States. The cancer workforce, which relies heavily on clinical teamwork and care coordination, may be more greatly impacted by turnover. In this study, we aimed to characterize oncologists who move to identify targets for recruitment and retention efforts. Methods: We identified medical, radiation, and surgical oncologists who treated Medicare beneficiaries diagnosed with breast, colorectal, or lung cancer in 2016–2019. We used multivariable logistic regression to identify physician-level and multivariable multinomial regression to identify region-level characteristics associated with turnover. Measures included demographic, practice, and patient-sharing network characteristics. Results: Our cohort included 25,012 medical, radiation, and surgical oncologists, of which, 1448 (5.8%) moved. Women [vs men; odds ratio (OR): 1.46; 95% CI: 1.30–1.64] and surgeons (vs medical oncologists; OR: 1.17; 95% CI; 1.04–1.33) had higher odds of moving. Compared with oncologists with moderate patient-sharing ties, those with many ties had lower odds of moving (OR: 0.55; 95% CI: 0.43–0.70). Patient-sharing networks with low efficiency (vs moderate) were more likely to have a net loss in their oncology workforce (OR: 3.06; 95% CI: 1.12–8.35), whereas those with low specialist vulnerability (vs moderate) were less likely to have a net loss (OR: 0.32; 95% CI: 0.1–0.99). Conclusions: This study identified novel patient-sharing network characteristics associated with turnover, providing new insights into how the structural features of patient-sharing networks may be related to the recruitment and retention of oncologists.

Variation in patient-sharing network characteristics of health care professionals treating different mental and substance use disorder patient sub-groups in primary care.

Providing efficient and targeted services for patients with mental health problems requires efficient collaboration and coordination within healthcare providers, but measuring collaboration using traditional methods is challenging. To explore the patient-sharing networks of professionals taking care of different groups of patients with mental or substance use disorders. We used data that covered adult patients' visits to the primary care service providers of seven municipalities in Finland during year 2021. Data included 8,217 patients (147,430 visits) with mental or substance use disorders who were treated by 1,566 health care professionals. We calculated descriptive network metrics to examine the connectivity of professionals in three different patient groups (patients with substance use disorders, psychotic disorders, and depressive disorders) and compared these characteristics to a network based on all patients. We also analyzed whether patient sharing was associated with the health care professionals' attributes (occupational group, municipality) using Exponential Random Graph Models (ERGM). Diagnosis-specific networks were denser and more connected compared to the all-patients network. Nurses were the most central occupation in all the diagnosis-specific networks and especially in the substance use disorder patients network. When examining all patients, two professionals were more likely to share patients when they belonged to the same occupational group. However, in the network with depressive disorder patients we found the opposite: professionals were more likely to share patients if they were of different occupational groups. Patient-sharing networks within patients with a specific mental or substance use disorders are denser and more connected than networks based on all patients with mental or substance use disorders. In the substance use disorder patients network particularly, nurses were the most central occupation. Multi-professional connections were more likely in depressive disorder networks than in the all-patients network.

Surgeon and Care Team Network Measures and Timely Breast Cancer Treatment

Cancer treatment delay is a recognized marker of worse outcomes. Timely treatment may be associated with physician patient-sharing network characteristics, yet this remains understudied. To examine the associations of surgeon and care team patient-sharing network measures with breast cancer treatment delay. This cross-sectional study of Medicare claims in a US population-based setting was conducted from 2017 to 2020. Eligible participants included patients with breast cancer who received surgery and the subset who went on to receive adjuvant therapy. Patient-sharing networks were constructed for treating physicians. Data were analyzed from September 2023 to February 2024. Surgeon linchpin score (a measure of local uniqueness or scarcity) and care density (a measure of physician team familiarity) were assessed. Surgeons were considered linchpins if their linchpin score was in the top 15%. The care density of a patient's physician team was calculated on preoperative teams for surgically-treated patients and postoperative teams for adjuvant therapy-receiving patients. The primary outcomes were surgical and adjuvant delay, which were defined as greater than 60 days between biopsy and surgery and greater than 60 days between surgery and adjuvant therapy, respectively. The study cohort included 56 433 patients (18 004 aged 70-74 years [31.9%]) who were mostly from urban areas (44 931 patients [79.6%]). Among these patients, 8009 (14.2%) experienced surgical delay. Linchpin surgeon status (locally unique surgeon) was not statistically associated with surgical delay; however, patients with high preoperative care density (ie, high team familiarity) had lower odds of surgical delay compared with those with low preoperative care density (odds ratio [OR], 0.58; 95% CI, 0.53-0.63). Of the 29 458 patients who received adjuvant therapy after surgery, 5700 (19.3%) experienced adjuvant delay. Patients with a linchpin surgeon had greater odds of adjuvant delay compared with those with a nonlinchpin surgeon (OR, 1.30; 95% CI, 1.13-1.49). Compared with those with low postoperative care density, there were lower odds of adjuvant delay for patients with high postoperative care density (OR, 0.77; 95% CI, 0.69-0.87) and medium postoperative care density (OR, 0.85; 95% CI, 0.77-0.94). In this cross-sectional study of Medicare claims, network measures capturing physician scarcity and team familiarity were associated with timely treatment. These results may help guide system-level interventions to reduce cancer treatment delays.

Exploiting relationship directionality to enhance statistical modeling of peer-influence across social networks.

Risky-prescribing is the excessive or inappropriate prescription of drugs that singly or in combination pose significant risks of adverse health outcomes. In the United States, prescribing of opioids and other "risky" drugs is a national public health concern. We use a novel data framework-a directed network connecting physicians who encounter the same patients in a sequence of visits-to investigate if risky-prescribing diffuses across physicians through a process of peer-influence. Using a shared-patient network of 10661 Ohio-based physicians constructed from Medicare claims data over 2014-2015, we extract information on the order in which patients encountered physicians to derive a directed patient-sharing network. This enables the novel decomposition of peer-effects of a medical practice such as risky-prescribing into directional (outbound and inbound) and bidirectional (mutual) relationship components. Using this framework, we develop models of peer-effects for contagion in risky-prescribing behavior as well as spillover effects. The latter is measured in terms of adverse health events suspected to be related to risky-prescribing in patients of peer-physicians. Estimated peer-effects were strongest when the patient-sharing relationship was mutual as opposed to directional. Using simulations we confirmed that our modeling and estimation strategies allows simultaneous estimation of each type of peer-effect (mutual and directional) with accuracy and precision. We also show that failing to account for these distinct mechanisms (a form of model mis-specification) produces misleading results, demonstrating the importance of retaining directional information in the construction of physician shared-patient networks. These findings suggest network-based interventions for reducing risky-prescribing.

Disparities in Access to Multidisciplinary Cancer Consultations and Treatment for Patients With Early-Stage Non-Small Cell Lung Cancer: A SEER-Medicare Analysis

PurposeDisparities in access to multidisciplinary cancer consultations (MDCc) persist, and the role of physician relationships remains understudied. This study examined the extent to which multilevel factors, including patient characteristics and patient-sharing network measures reflecting the structure of physician relationships, are associated with MDCc and receipt of stereotactic body radiation therapy (SBRT) versus surgery among early-stage non-small cell lung cancer (NSCLC) patients. Materials and MethodsIn this cross-sectional study, we analyzed Surveillance, Epidemiology, and End-Result (SEER)-Medicare data for patients diagnosed with stage I-IIA NSCLC in 2016-2017. We assembled patient-sharing networks and identified cancer specialists who were locally unique for their specialty, herein referred to as “linchpins”. The proportion of linchpin cancer specialists for each hospital referral region (HRR) was calculated as a network-based measure of specialist scarcity. We used multilevel multinomial logistic regression to estimate associations between study variables and the receipt of MDCc and multilevel logistic regression to examine the relationship between MDCc and patient's first treatment. ResultsOur study included 6,120 patients with stage I-IIA NSCLC, of which 751 (12.3%) received MDCc, 1,729 (28.3%) only consulted a radiation oncologist, 2,010 (32.8%) only consulted a surgeon, and 1,630 (26.6%) had no consultations with either specialist within two months following diagnosis. Compared with patients residing in an HRR with a low proportion of linchpin surgeons, those in an HRR with a high proportion of linchpin surgeons had a 2.99 (95% CI: 1.87-4.78) greater relative risk of exclusively consulting a radiation oncologist (vs. MDCc) and a 2.70 (95% CI: 1.68-4.35) greater relative risk consulting neither specialist (vs. MDCc). Patients who received MDCc were 5.32 (95% CI: 4.27-6.63) times more likely to receive SBRT (vs. surgery). ConclusionsPhysician networks are associated with receipt of MDCc and treatment, underscoring the potential for leveraging patient-sharing network analysis to improve access to lung cancer care.

Patient-sharing relationships among primary healthcare professionals taking care of patients with mental health problems or substance abuse: the social network approach.

Introduction: Mental health disorders are recognized as considerable public health challenges that require close collaboration between professionals. Developing collaboration requires understanding the structure of how patient care is managed across professionals. Healthcare registers offer “naturally occuring” data of how different professionals are involved in patient care.
 Research question: To find out which local and individual factors influence patient-sharing network structure and composition.
 Methodologies: Data were from the Register of Primary Health Care visits (Avohilmo) which covers all outpatient primary healthcare visits in Finland. We combined data on visits from eight health and social centres during the year 2021. We further included to the data only those clients’ visits, who had at least one visit relating primarily to mental health or substance abuse. The patient-sharing networks were formed using the unique ID codes of the professionals and different occupational groups included in the register. Two professionals were considered to have a patient-sharing relationship if they both had delivered care to the same patient. The resulting network was a one-mode undirected network graph with 646 nodes (healthcare professionals) and 50 841 unweighted ties (the number of shared patients) between them. These ties were based on 11 796 patients and 126 521 visits.
 Statistical analyses: We analyzed the potential associations of the network structure and the nodal attributes (working in a specific health and social centre, belonging to a certain occupational group and the reason of the patient visit) with nodal formation using Exponential Random Graph Models (ERGM). ERGMs are statistical models that enable network pattern recognition and are similar to logistic regression: they predict the probability for a pair of nodes in a network to have a tie between them (in this case that two professionals share the same patient).
 Results: Our results show that there are uniform effects regarding occupational group and health and social centre. Two professionals were more likely to share a patient if they had similar occupation and naturally when they worked in the same centre. The coefficient for shared partners was positive indicating a propensity for a triadic closure in the network. The main effects were also significant suggesting that being a physician predicts having more connections than belonging to other occupational groups. There were also significant differences between the centres suggesting that the differences in degrees between centres were not random. More importantly, the main effect of the reason for the patient visit was significant, suggesting that when the reason for visit was related to mental health problems or substance abuse the connections with other professionals were more probable, than when the reason was related to other health issues. 
 Conclusion: The results suggest that professionals seem to work more often within their own occupational group than with other professionals. Contacts between professionals were more probable when the reason for visit was related to mental health problems or substance abuse. Further research is needed to examine, whether sharing patients also indicate that there is collaboration between the professionals.

Do Physician Patient Sharing Networks Impact Disparities in Access to Multidisciplinary Cancer Consultation for Early-Stage Lung Cancer? A SEER-Medicare Network Analysis

To examine early-stage non-small cell lung cancer (NSCLC) patient and physician network characteristics associated with receipt of multidisciplinary cancer consultations (MDCCs), defined as an encounter with both a surgeon and a radiation oncologist after diagnosis. We tested the hypothesis that physician roles in patient-sharing networks can modify disparities in access to surgical and radiotherapy care. Specifically, we investigated the role of "linchpin physicians" using a novel network measure that assesses the extent to which a physician is the sole type of specialist in their physician patient sharing network, and the association of these physicians on patterns of care. The study cohort included patients diagnosed with stage I/IIA NSCLC in 2016-17 using Surveillance, Epidemiology and End Results (SEER)-Medicare data. We assembled a physician patient-sharing network using all NSCLC patient encounters between 3 months prior to and 12 months following diagnosis. We weighted physicians' patient-sharing ties by the number of shared patients and calculated physician's linchpin scores as a continuous variable measuring the extent to which a physicians' peers in the network lack ties to others of the same specialty as the focal physician. Physicians in the top 15%ile were considered a linchpin specialist. We then derived the proportion of linchpin surgeons or radiation oncologists per hospital referral region (HRR). We assessed patient characteristics including age, sex, race, ethnicity, cancer stage, comorbidity index, rurality, and socioeconomic status (SES). The study outcome was a surgical, radiation oncology, or MDCC in the 2 months following diagnosis. We then performed multivariable logistic regressions (MLR) to assess whether linchpins modified disparities in MDCC. Of 6,148 patients in the study, 12% received a MDCC, 28% only saw a radiation oncologist, 33% only saw a surgeon, and 27% saw neither for 2 months after diagnosis. On MLR, Hispanic patients were significantly less likely than White patients to receive consultations with a radiation oncologist, while patients in lowest SES quintile were significantly less likely to receive a surgical consultation than those in the highest SES. The proportion of linchpin surgeons and radiation oncologists per HRR was negatively associated with receipt of MDCC (OR low vs high [95% CI] = 0.75 [0.66 to 0.86]; 0.70 [0.61 to 0.80, respectively), but our regressions did not show that any demographic or SES disparities were modified by linchpin specialists. Linchpin specialists do not impact known disparities in access to cancer consultations. Identifying additional system-level factors that contribute to disparities in cancer consultations may lead to policy insights aimed at reducing unwarranted variation in care.

An interactive patient transfer network and model visualization tool for multidrug-resistant organism prevention strategies

Background: The CDC’s new Public Health Strategies to Prevent the Spread of Novel and Targeted Multidrug-Resistant Organisms (MDROs) were informed by mathematical models that assessed the impact of implementing preventive strategies directed at a subset of healthcare facilities characterized as influential or highly connected based on their predicted role in the regional spread of MDROs. We developed an interactive tool to communicate mathematical modeling results and visualize the regional patient transfer network for public health departments and healthcare facilities to assist in planning and implementing prevention strategies. Methods: An interactive RShiny application is currently hosted in the CDC network and is accessible to external partners through the Secure Access Management Services (SAMS). Patient transfer volumes (direct and indirect, that is, with up to 30 days in the community between admissions) were estimated from the CMS fee-for-service claims data from 2019. The spread of a carbapenem-resistant Enterobacterales (CRE)–like MDROs within a US state was simulated using a deterministic model with susceptible and infectious compartments in the community and healthcare facilities interconnected through patient transfers. Individuals determined to be infectious through admission screening, point-prevalence surveys (PPSs), or notified from interfacility communication were assigned lower transmissibility if enhanced infection prevention and control practices were in place at a facility. Results: The application consists of 4 interactive tabs. Users can visualize the statewide patient-sharing network for any US state and select territories in the first tab (Fig. 1). A feature allows users to highlight a facility of interest and display downstream or upstream facilities that received or sent transfers from the facility of interest, respectively. A second tab lists influential facilities to aid in prioritizing screening and prevention activities. A third tab lists all facilities in the state in descending order of their dispersal rate (ie, the rate at which patients are shared downstream to other facilities), which can help identify highly connected facilities. In the fourth tab, an interactive graph displays the predicted reduction of MDRO prevalence given a range of intervention scenarios (Fig. 2). Conclusions: Our RShiny application, which can be accessed by public health partners, can assist healthcare facilities and public health departments in planning and tailoring MDRO prevention activity bundles.Disclosures: None

Use of a Novel Network-Based Linchpin Score to Characterize Accessibility to the Oncology Physician Workforce in the United States

Physician headcounts provide useful information about the cancer care delivery workforce; however, efforts to track the oncology workforce would benefit from new measures that capture how essential a physician is for meeting the multidisciplinary cancer care needs of the region. Physicians are considered linchpins when fewer of their peers are connected to other physicians of the same specialty as the focal physician. Because they are locally unique for their specialty, these physicians' networks may be particularly vulnerable to their removal from the network (eg, through relocation or retirement). To examine a novel network-based physician linchpin score within nationwide cancer patient-sharing networks and explore variation in network vulnerability across hospital referral regions (HRRs). This cross-sectional study analyzed fee-for-service Medicare claims and included Medicare beneficiaries with an incident diagnosis of breast, colorectal, or lung cancer from 2016 to 2018 and their treating physicians. Data were analyzed from March 2022 to October 2022. Physician characteristics assessed were specialty, rurality, and Census region. HRR variables assessed include sociodemographic and socioeconomic characteristics and use of cancer services. Oncologist linchpin score, which examined the extent to which a physician's peers were connected to other physicians of the same specialty as the focal physician. Network vulnerability, which distinguished HRRs with more linchpin oncologists than expected based on oncologist density. χ2 and Fisher exact tests were used to examine relationships between oncologist characteristics and linchpin score. Spearman rank correlation coefficient (ρ) was used to measure the strength and direction of relationships between HRR network vulnerability, oncologist density, population sociodemographic and socioeconomic characteristics, and cancer service use. The study cohort comprised 308 714 patients with breast, colorectal, or lung cancer. The study cohort of 308 714 patients included 161 206 (52.2%) patients with breast cancer, 76 604 (24.8%) patients with colorectal cancer, and 70 904 (23.0%) patients with lung cancer. In our sample, 272 425 patients (88%) were White, and 238 603 patients (77%) lived in metropolitan areas. The cancer patient-sharing network included 7221 medical oncologists and 3573 radiation oncologists. HRRs with more vulnerable networks for medical oncology had a higher percentage of beneficiaries eligible for Medicaid (ρ, 0.19; 95% CI, 0.08 to 0.29). HRRs with more vulnerable networks for radiation oncology had a higher percentage of beneficiaries living in poverty (ρ, 0.17; 95% CI, 0.06 to 0.27), and a higher percentage of beneficiaries eligible for Medicaid (ρ, 0.21; 95% CI, 0.09 to 0.31), and lower rates of cohort patients receiving radiation therapy (ρ, -0.18; 95% CI, -0.28 to -0.06; P = .003). The was no association between network vulnerability for medical oncology and percent of cohort patients receiving chemotherapy (ρ, -0.03; 95% CI, -0.15 to 0.08). This study found that patient-sharing network vulnerability was associated with poverty and lower rates of radiation therapy. Health policy strategies for addressing network vulnerability may improve access to interdisciplinary care and reduce treatment disparities.

The influence of strong and weak ties in physician peer networks on new drug adoption.

Physicians interact and exchange information through various social networks. Understanding peer effects through different networks can help accelerate new medical technology and innovative treatment adoption. In this research, we measure the influence of strong-tie and weak-tie connections on new drug adoption and study the overlap between advice-discussion and patient-sharing network. We construct two physician networks with strong and weak ties from peer nomination surveys and commercial medical claims data. We design a dynamic system to define peer adoption status and build patient-level hierarchical logistic models to measure the peer influence on new product adoption for treating new-to-therapy patients. Our results show that A strong-tie early adoption peer has six times more influence on new drug adoption than a weak-tie peer. Weak tie peers collectively exert as much or higher influence than strong-tie peers because of the larger network size. In the case of inaccessibility to strong-tie data, researchers can still reliably use the influence of the weak tie data only even though they will lose the effect of the omitted strong ties.

Beyond patient-sharing: Comparing physician- and patient-induced networks

The sharing of patients reflects collaborative relationships between various healthcare providers. Patient-sharing in the outpatient sector is influenced by both physicians’ activities and patients’ preferences. Consequently, a patient-sharing network arises from two distinct mechanisms: the initiative of the physicians on the one hand, and that of the patients on the other. We draw upon medical claims data to study the structure of one patient-sharing network by differentiating between these two mechanisms. Owing to the institutional requirements of certain healthcare systems rather following the Bismarck model, we explore different triadic patterns between general practitioners and medical specialists by applying exponential random graph models. Our findings imply deviation from institutional expectations and reveal structural realities visible in both networks.

The association of insurance plan characteristics with physician patient-sharing network structure.

Professional and social connections among physicians impact patient outcomes, but little is known about how characteristics of insurance plans are associated with physician patient-sharing network structure. We use information from commercially insured enrollees in the 2011 Massachusetts All Payer Claims Database to construct and examine the structure of the physician patient-sharing network using standard and novel social network measures. Using regression analysis, we examine the association of physician patient-sharing network measures with an indicator of whether a patient is enrolled in a health maintenance organization (HMO) or preferred provider organization (PPO), controlling for patient and insurer characteristics and observed health status. We find patients enrolled in HMOs see physicians who are more central and densely embedded in the patient-sharing network. We find HMO patients see PCPs who refer to specialists who are less globally central, even as these specialists are more locally central. Our analysis shows there are small but significant differences in physician patient-sharing network as experienced by patients with HMO versus PPO insurance. Understanding connections between physicians is essential and, similar to previous findings, our results suggest policy choices in the insurance and delivery system that change physician connectivity may have important implications for healthcare delivery, utilization and costs.

Dispersion in the hospital network of shared patients is associated with less efficient care.

There is growing recognition that health care providers are embedded in networks formed by the movement of patients between providers. However, the structure of such networks and its impact on health care are poorly understood. We examined the level of dispersion of patient-sharing networks across U.S. hospitals and its association with three measures of care delivered by hospitals that were likely to relate to coordination. We used data derived from 2016 Medicare Fee-for-Service claims to measure the volume of patients that hospitals treated in common. We then calculated a measure of dispersion for each hospital based on how those patients were concentrated in outside hospitals. Using this measure, we created multivariate regression models to estimate the relationship between network dispersion, Medicare spending per beneficiary, readmission rates, and emergency department (ED) throughput rates. In multivariate analysis, we found that hospitals with more dispersed networks (those with many low-volume patient-sharing relationships) had higher spending but not greater readmission rates or slower ED throughput. Among hospitals with fewer resources, greater dispersion related to greater readmission rates and slower ED throughput. Holding an individual hospital's dispersion constant, the level of dispersion of other hospitals in the hospital's network was also related to these outcomes. Dispersed interhospital networks pose a challenge to coordination for patients who are treated at multiple hospitals. These findings indicate that the patient-sharing network structure may be an overlooked factor that shapes how health care organizations deliver care. Hospital leaders and hospital-based clinicians should consider how the structure of relationships with other hospitals influences the coordination of patient care. Effective management of this broad network may lead to important strategic partnerships.

Estimating the Impact of County Boundaries on State-wide Patient-Sharing Network Models

Background: In the field of public health, network models are useful for understanding the spread of both information and infectious diseases. Collecting network data requires determining network boundaries (ie, the entities selected for data collection). These decisions, if not made carefully, have potential outsized downstream effects on study findings. In practice, collaboration and coordination between healthcare organizations are often dictated by historical or geopolitical boundaries (eg, state or county boundaries), which may distort the underlying network under study, and thereby affect the reliability and/or accuracy of the network model. Objective: We compared natural communities in a patient-sharing network with those induced by geopolitical boundaries. Methods: Using data from the Healthcare Cost and Utilization Project (HCUP), we constructed a patient-sharing network among hospitals in California, splitting the data into a training set and a holdout set. We performed edge-betweenness clustering on the training set, and with the holdout set, we compared the resulting partition with the partition by counties using modularity. We also clustered contiguous counties that might function more cohesively together than individually. We performed spatially constrained hierarchical clustering on the network constructed from total patient flow between pairs of counties. The results were again compared via modularity on the holdout set to the county partition. Lastly, we built an individual-based model (IBM) using HCUP and American Hospital Association data to perform epidemic simulations. For each of several counties, we implemented this model to estimate the proportion of patients infected over time. We then reran the individual-based model using the entire state while dividing the results into corresponding counties. Results: In total, 680,485 patients transferred between 374 hospitals in 55 counties from 2003 to 2011. The out-of-sample modularity for the edge-betweenness clustering partition was 464% higher than that of the county partition. Aggregating the counties into half as many contiguous clusters was 319% higher, and aggregating them into 6 clusters was 489% higher (Fig. 1). The epicurves from the individual-based model ranged from small to significant deviations between state versus county boundaries (Fig. 2) . Conclusions: Collecting network data using externally imposed boundaries may lead to inaccurate network models. For example, counties serve as a poor proxy for their underlying communities, resulting in poor overall disease spread simulation results when county boundaries are allowed to drive network construction. These issues should be considered when building coordination partnerships such as the Accountable Communities for Health.Funding: NoneDisclosures: None

Modeling peer effect modification by network strength: The diffusion of implantable cardioverter defibrillators in the US hospital network.

We develop methodology that allows peer effects (also referred to as social influence and contagion) to be modified by the structural importance of the focal actor's position in the network. The methodology is first developed for a single peer effect and then extended to simultaneously model multiple peer-effects and their modifications by the structural importance of the focal actor. This work is motivated by the diffusion of implantable cardioverter defibrillators (ICDs) in patients with congestive heart failure across a cardiovascular disease patient-sharing network of United States hospitals. We apply the general methodology to estimate peer effects for the adoption of capability to implant ICDs, the number of ICD implants performed by hospitals that are capable, and the number of patients referred to other hospitals by noncapable hospitals. Applying our novel methodology to study ICD diffusion across hospitals, we find evidence that exposure to ICD-capable peer hospitals is strongly associated with the chance a hospital becomes ICD-capable and that the direction and magnitude of the association is extensively modified by the strength of that hospital's position in the network, even after controlling for effects of geography. Therefore, interhospital networks, rather than geography per se, may explain key patterns of regional variations in healthcare utilization.

Estimating the Attributable Disease Burden and Effects of Interhospital Patient Sharing on Clostridium difficile Infections.

To estimate the burden of Clostridium difficile infections (CDIs) due to interfacility patient sharing at regional and hospital levels. Retrospective observational study. We used data from the Healthcare Cost and Utilization Project California State Inpatient Database (2005-2011) to identify 26,878,498 admissions and 532,925 patient transfers. We constructed a weighted, directed network among the hospitals by defining an edge between 2 hospitals to be the monthly average number of patients discharged from one hospital and admitted to another on the same day. We then used a network autocorrelation model to study the effect of the patient sharing network on the monthly average number of CDI cases per hospital, and we estimated the proportion of CDI cases attributable to the network. We found that 13% (95% confidence interval [CI], 7.6%-18%) of CDI cases were due to diffusion through the patient-sharing network. The network autocorrelation parameter was estimated at 5.0 (95% CI, 3.0-6.9). An increase in the number of patients transferred into and/or an increased CDI rate at the hospitals from which those patients originated led to an increase in the number of CDIs in the receiving hospital. A minority but substantial burden of CDI infections are attributable to hospital transfers. A hospital's infection control may thus be nontrivially influenced by its neighboring hospitals. This work adds to the growing body of evidence that intervention strategies designed to minimize HAIs should be done at the regional rather than local level.

Influence of peer networks on physician adoption of new drugs.

Although physicians learn about new medical technologies from their peers, the magnitude and source of peer influence is unknown. We estimate the effect of peer adoption of three first-in-class medications (dabigatran, sitigliptin, and aliskiren) on physicians’ own adoption of those medications. We included 11,958 physicians in Pennsylvania prescribing anticoagulant, antidiabetic, and antihypertensive medications. We constructed 4 types of peer networks based on shared Medicare and Medicaid patients, medical group affiliation, hospital affiliation, and medical school/residency training. Instrumental variables analysis was used to estimate the causal effect of peer adoption (fraction of peers in each network adopting the new drug) on physician adoption (prescribing at least the median number prescriptions within 15 months of the new drug’s introduction). We illustrate how physician network position can inform targeting of interventions to physicians by computing a social multiplier. Dabigatran was adopted by 25.2%, sitagliptin by 24.5% and aliskiren by 8.3% of physicians. A 10-percentage point increase in peer adoption in the patient-sharing network led to a 5.90% (SE = 1.50%, p<0.001) increase in physician adoption of dabigatran, 8.32% (SE = 1.51%, p<0.001) increase in sitagliptin, and 7.84% increase in aliskiren adoption (SE = 2.93%, p<0.001). Peer effects through shared hospital affiliation were positive but not significant, and medical group and training network effects were not reliably estimated. Physicians in the top decile of patient-sharing network peers were estimated to have nearly 2-fold stronger influence on their peers’ adoption compared to physicians in the top decile of prescribing volume. Limitations include lack of detailed clinical information and pharmaceutical promotion, variables which may influence physician adoption but which are unlikely to bias our peer effect estimates. Peer adoption, especially by those with whom physicians share patients, strongly influenced physician adoption of new drugs. Our study shows the potential for using information on physician peer networks to improve technology diffusion.

The nature and scope of patient-sharing network research: a novel, important area for network science.

A robust literature exists that draws on social network approaches to understand connections among individuals, and healthcare and behavior-related implications. This article offers commentary on the scoping review conducted by Dugoff et al. that examines "patient-sharing" networks, their characteristics, and various methodological approaches. The scoping review conducted by Dugoff et al. examines the characteristics and methods employed in patient-sharing network studies. It identified the most common measures used in patient-sharing network research, as well as theories used in patient-sharing network studies. Dugoff et al. also identified many studies that examined healthcare utilization considerations for patient sharing. Understanding the connections between providers and the flow of patients between providers could lend insight into barriers and enablers to efficient healthcare systems.

A scoping review of patient-sharing network studies using administrative data.

There is a robust literature examining social networks and health, which draws on the network traditions in sociology and statistics. However, the application of social network approaches to understand the organization of health care is less well understood. The objective of this work was to examine approaches to conceptualizing, measuring, and analyzing provider patient-sharing networks. These networks are constructed using administrative data in which pairs of physicians are considered connected if they both deliver care to the same patient. A scoping review of English language peer-reviewed articles in PubMed and Embase was conducted from inception to June 2017. Two reviewers evaluated article eligibility based upon inclusion criteria and abstracted relevant data into a database. The literature search identified 10,855 titles, of which 63 full-text articles were examined. Nine additional papers identified by reviewing article references and authors were examined. Of the 49 papers that met criteria for study inclusion, 39 used a cross-sectional study design, 6 used a cohort design, and 4 were longitudinal. We found that studies most commonly theorized that networks reflected aspects of collaboration or coordination. Less commonly, studies drew on the strength of weak ties or diffusion of innovation frameworks. A total of 180 social network measures were used to describe the networks of individual providers, provider pairs and triads, the network as a whole, and patients. The literature on patient-sharing relationships between providers is marked by a diversity of measures and approaches. We highlight key considerations in network identification including the definition of network ties, setting geographic boundaries, and identifying clusters of providers, and discuss gaps for future study.

Spread of Carbapenem-Resistant Enterobacteriaceae Among Illinois Healthcare Facilities: The Role of Patient Sharing.

Carbapenem-resistant Enterobacteriaceae (CRE) spread regionally throughout healthcare facilities through patient transfer and cause difficult-to-treat infections. We developed a state-wide patient-sharing matrix and applied social network analyses to determine whether greater connectedness (centrality) to other healthcare facilities and greater patient sharing with long-term acute care hospitals (LTACHs) predicted higher facility CRE rates. We combined CRE case information from the Illinois extensively drug-resistant organism registry with measures of centrality calculated from a state-wide hospital discharge dataset to predict facility-level CRE rates, adjusting for hospital size and geographic characteristics. Higher CRE rates were observed among facilities with greater patient sharing, as measured by degree centrality. Each additional hospital connection (unit of degree) conferred a 6% increase in CRE rate in rural facilities (relative risk [RR] = 1.056; 95% confidence interval [CI], 1.030-1.082) and a 3% increase among Chicagoland and non-Chicago urban facilities (RR = 1.027; 95% CI, 1.002-1.052 and RR = 1.025; 95% CI, 1.002-1.048, respectively). Sharing 4 or more patients with LTACHs was associated with higher CRE rates, but this association may have been due to chance (RR = 2.08; 95% CI, .85-5.08; P = .11). Hospitals with greater connectedness to other hospitals in a statewide patient-sharing network had higher CRE burden. Centrality had a greater effect on CRE rates in rural counties, which do not have LTACHs. Social network analysis likely identifies hospitals at higher risk of CRE exposure, enabling focused clinical and public health interventions.