Institutional review board approval was obtained. Informed consent and assent, when appropriate, were obtained from participants at the time of enrollment.
From March 2010 to March 2011, patients were recruited at a county hospital sports medicine clinic and a university-based sports medicine practice in South Florida. The inclusion criteria allowed for the enrollment of any patient with an ACL tear that had been diagnosed clinically and confirmed with magnetic resonance imaging (MRI). Patients with multiligamentous knee injuries and polytrauma were excluded. Eighty-one consecutive subjects who met the selection criteria were identified for participation. One subject refused, yielding an enrollment rate of 98.8%. The date of the injury was considered to be time zero. Diagnosis of an ACL tear, confirmed by means of physical examination and MRI by one of two sports medicine fellowship-trained orthopaedic surgeons, was the end point. Data were collected by means of a patient interview in the outpatient clinical setting.
Demographic data were collected on age, sex, and employment status. Insurance status was classified as uninsured, government, or private. The dates of the injury and the first healthcare encounter (classified as either emergency department or primary-care physician/orthopaedic surgeon) were documented. The total number of visits was calculated, including a visit for MRI evaluation if performed on a separate date than an office visit. To minimize patient recall as the only source of information, information obtained from patient interviews was corroborated through electronic medical records when available. The total time from the injury to the initial presentation and the time to the definitive diagnosis were calculated. Time was classified as patient time or system time. Patient time encompassed any voluntary reason for not seeking care. System time encompassed any reason that could be attributed to either the insurance or healthcare system, such as waiting for the next available appointment, lack of referral to a specialist, or an inability to find a specialist who accepted the patient’s insurance. An adjusted time to diagnosis was calculated by subtracting patient time from the total time to diagnosis. From the records documenting the date when the MRI was ordered, the date when it was performed, and the date of the follow-up visit, we were able to calculate the time interval between each time point.
Statistical Analysis
Patients were divided into three groups based on insurance status: (1) uninsured, (2) government insurance (Medicaid), and (3) private insurance. Categorical data and time outcome variables were compared among the groups with use of the Kruskal-Wallis test or the Fisher exact test as appropriate. Differences among the three groups in terms of the total time to diagnosis and the adjusted time to diagnosis were analyzed with the Wilcoxon-Mann-Whitney test. With use of Cox proportional-hazards regression, we calculated crude and adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) to determine the effect of insurance status on the time to diagnosis of the ACL tear; the government insurance group was used as the referent group. Our models controlled for the variables of age, sex, type of first visit, and employment status. For all statistical analyses, the level of significance was defined as p < 0.05. Statistical analyses were conducted with use of SAS software (version 9.2; SAS Institute, Cary, North Carolina) and STATA (version 11.1; StataCorp, College Station, Texas).
Source of Funding
There were no external sources of funding for this study.
Eighty patients were included in the study, including twelve uninsured patients, twenty patients with government insurance, and forty-eight patients with private insurance. Patient characteristics by insurance type can be seen in Table I. There were significant differences among the groups with regard to age and employment status (p < 0.05). There was no difference among the three groups with regard to sex or the median time to the first visit. The type of first visit differed among the three groups, with more uninsured patients (nine of twelve) and government-insured patients (fourteen of twenty) visiting the emergency department for their initial visit than those with private insurance (twenty of forty-eight) (p = 0.032). A significant difference in the total number of visits to diagnosis was also seen, with uninsured and government-insured patients requiring a median of four visits as compared with three visits for privately insured patients (p = 0.006).
Time Outcome Variables
Significant differences were seen in time outcome variables, including total time to diagnosis, adjusted time to diagnosis, system time to diagnosis, time to obtain MRI, time to follow-up after MRI, and total time from the ordering of MRI to follow-up (p ≤ 0.002) (Table I). The median total times from the injury to the diagnosis of ACL tear were 121 days (range, forty-five to 437 days), fifty-six days (range, twelve to 402 days), and fourteen days (range, two to 469 days) for the uninsured, government-insured, and privately insured patients, respectively (p < 0.001). When patient time was subtracted, the diagnosis was made at a median of eighty-three days (range, forty-five to 434 days) for the uninsured group, forty-three days (range, four to 388 days) for the government insurance group, and eight days (range, one to 223 days) for the private insurance group. No significant difference was observed among the groups in terms of patient time. The median time between the ordering of MRI and the performance of the MRI was thirty days (range, seventeen to 107 days) for the uninsured group, eight days (range, zero to forty-five days) for the government insurance group, and three days (range, zero to twelve days) for the private insurance group (p < 0.001). When we combined the time from when the MRI was ordered and obtained to the time for the patient to return to the office for follow-up and diagnosis, the median time was sixty-four days (range, thirty-five to 227 days) for the uninsured group, ten days (range, four to ninety-three days) for the government insurance group, and six days (range, zero to fifteen days) for the private insurance group (p < 0.001). Patients who were seen initially by their primary-care physicians or an orthopaedic surgeon were diagnosed at a median of ten days (range, two to 434 days), whereas those who visited the emergency department were diagnosed at a median of twenty-nine days (range, one to 388 days) (p = 0.016), regardless of insurance type.
Total and Adjusted Time to Diagnosis by Insurance Type
The total time to diagnosis and the adjusted time to diagnosis were analyzed individually for differences among the three groups (Table II). Significant differences were found between the private insurance group and both the uninsured and government insurance groups with regard to both the total time to diagnosis and the adjusted time to diagnosis (p < 0.001). No difference was observed between the uninsured and government insurance groups with regard to the total time to diagnosis or the adjusted time to diagnosis. After controlling for age, sex, type of first visit, and employment status, patients with private insurance were 2.15 times (95% CI, 1.18 to 3.93 times) (p = 0.012) and 4.53 times (95% CI, 2.43 to 8.42 times) (p < 0.001) more likely to be diagnosed in a given time period than those with government insurance in terms of the total time to diagnosis and the adjusted time to diagnosis, respectively (Table III). There was no significant difference between the uninsured group and the government insurance group.
Our study supported our hypothesis that, when grouped by insurance type, patients receiving Medicaid and those with no insurance face significantly greater delays in obtaining care in comparison with those with private insurance in South Florida. Although our results reflect a single geographic location, the findings of this study are consistent with those of previous regional and national investigations involving orthopaedics and access to care. Skaggs et al.5,6 previously documented this disparity at both the state and national levels in the orthopaedic literature. By contacting orthopaedic surgeons in the community with a specific scenario of a child with a fracture and only changing the scenario by the child’s insurance status (Medicaid or private), they were able to highlight the difficulties of obtaining care for patients with Medicaid insurance. Hinman and Bozic26 reported on the differences among patients undergoing primary total hip arthroplasty according to insurance type. In their series, patients with Medicaid had to travel twice as far to receive treatment and had significantly lower preoperative and postoperative Harris hip scores than did patients with Medicare or private insurance. The authors hypothesized that the decreased access to care and the increased time to presentation directly affected the functional results and clinical outcomes.
In the current study, several significant differences were found among the groups. The difference in the time from injury to diagnosis was markedly different between patients with private insurance and those with no insurance or government insurance. A significant difference in time to diagnosis was seen between the uninsured and government insurance groups only after taking into account patients who did not seek care initially. This finding could be explained by patient time representing a greater delay for the government insurance group as compared with the uninsured group, although this difference was not significant.
Patients in the private insurance group had a greater than twofold likelihood of being diagnosed in a given time period in comparison with those in the government insurance group. The likelihood of being diagnosed more than doubled when patient-related delays were considered, while no difference was seen between the uninsured group and the government insurance group in either analysis. This finding suggests that when patient-related delays are adjusted for, the discrepancies in access to care become more evident. Despite this finding, the difference in patient time was not significant between the groups.
The present study considered patient time as a delay in seeking initial care. Further delays in seeking care could be secondary to missed appointments, not seeking further care for economic reasons, inadequate transportation, or the inability to obtain time off from work. Children face additional barriers, as they depend on adults to be available to provide transportation and to provide economically for their care. Although the calculation of adjusted time to diagnosis attempted to account for these patient-related delays in seeking care, it was not possible to include such detail in the entire analysis. Our subjects provided numerous reasons for delays in seeking or obtaining care, which were classified as patient time. Other reasons for not obtaining care (including misdiagnosis, incorrect referral, time waiting for the next available appointment, referral from another provider, or inability to locate a provider who accepted their insurance) were classified as system time. The reasons for not seeking or obtaining care are more complex than can be classified into one of two categories, so care must be taken when considering the results as strictly patient or system delay.
Additional data regarding all subjects with government insurance are available in the Appendix.
Several additional factors play a role in increasing the time to diagnosis. Appointment availability is important to consider, as are the resources available at the treating institutions. Allocated clinic time at the county hospital study site was one half day per week, allowing for limited appointment slots. Although delays due to lack of available appointments are not directly due to insurance status, these delays are intimately related to insurance status. A lack of available providers accepting Medicaid would result in patients seeking care at or being referred to a county institution and thus being subject to these delays, whereas those with private insurance could choose to go elsewhere.
Patients in the uninsured and government insurance groups required more visits prior to diagnosis. This finding could be related to the greater number of patients in the uninsured and government insurance groups who sought initial care at the emergency department as compared with the number who did so in the private insurance group. The use of the emergency department for non-emergent conditions has been linked to insurance status27 and has been shown to increase healthcare costs28. Visits to the emergency department as well as the difficulty of obtaining follow-up visits with an orthopaedic surgeon play an important role in delaying the diagnosis. Sabharwal et al.29 studied pediatric orthopaedic patients who presented to a university emergency department after visiting another emergency department. Those with no insurance or government insurance were more likely than those with private insurance to have visited an additional healthcare facility before visiting the initial emergency department. Attempts to address this disparity by establishing a Medicaid managed-care system resulted in an eightfold increase in return visits to the emergency department for pediatric orthopaedic patients due to the inability to obtain follow-up30.
A difference in age was seen among the groups, with the Medicaid group having a significantly lower median age (seventeen years) than the uninsured and private insurance groups. This difference could be explained by the eligibility requirements for Medicaid. Although a pediatric patient population faces additional barriers to obtaining care3, we believe that this remains an important analysis to include in the study as numerous studies have highlighted increased meniscal and chondral injuries in association with a longer time from injury to treatment in this population18,21.
The purpose of including only patients with isolated ACL injuries was to use this diagnosis as a vehicle to analyze discrepancies in access to care, given the target age group to be analyzed and the non-emergent nature of the diagnosis. We theorized that the injury would be acute enough for the patients to seek care at a reasonable time interval and that the inability of the patients to continue the activities through which they had sustained the injury would drive them to seek care in an urgent yet non-emergent manner, providing an appropriate scenario to study access to care.
The present study had several weaknesses. Recall bias must be considered, as some patients had sustained the injury more than one year before enrolling in the study. Data were collected by means of a direct interview at the time of enrollment into the study, providing more accurate and study-specific information than a retrospective chart review. When available, patient history was corroborated through electronic medical records to help reduce recall bias. We included the time from the ordering of an MRI to the performance of the MRI as well as the time from the performance of the MRI to the follow-up appointment after the MRI. This information was recorded only when a progress note in the chart documented the information and the date of the imaging study, meaning data were not available for all patients. By using only documented dates of MRI orders and image acquisition, recall bias was eliminated from this data set. The majority of these data were obtained from patients who had imaging studies that were ordered by the treating orthopaedic surgeon, suggesting that the principal point of delay could be access to a specialist.
Another weakness is the small sample of uninsured patients. Although a significant difference in the time to diagnosis was found between the private insurance group and both the Medicaid and uninsured groups, no such difference was observed between the Medicaid and uninsured groups. The lack of significance between these two groups should be interpreted as a lack of statistical power rather than a lack of difference, given the small sample size of the uninsured population in our study. Although the determination that patients with Medicaid face fewer delays than those without insurance is important, we believe that the main purpose of the present study was to compare patients receiving Medicaid with those who had private insurance. As established by the Social Security Act, Section 1902(a)(30)(A), the goal of Medicaid as enforced by each state should be to ensure that “care and services are available under the plan at least to the extent that such care and services are available to the general population in the geographic area.” Our study shows that there may be a disparity in the availability or access to these services.
Another weakness that must be highlighted is the inclusion of only one geographic location in an urban setting, which may not be representative of other locations. Hospital-specific factors such as a lack of available appointments or resources can influence the results and may not hold true across all hospitals and geographic locations. Although our results are consistent with those of similar geographically limited studies, further research with multiple centers and locations would better serve to analyze differences at the national level. Another useful future study would investigate the long-term outcomes in these three groups to determine if the delay in diagnosis leads to a difference in functional outcome.
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. None of the authors, or their institution(s), have had any financial relationship, in the thirty-six months prior to submission of this work, with any entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, no author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.