Patterns of Primary Care Use Prior to a First Diagnosis of Nonaffective Psychotic Disorder in Ontario,Canada: Modèles d’utilisation des soins de première ligne avant un premier diagnostic de trouble psychotique non affectif en Ontario,Canada

Abstract

French

Background:

Many people experience early signs and symptoms before the onset of psychotic disorder, suggesting that there may be help-seeking prior to first diagnosis. The family physician has been found to play a key role in pathways to care. This study examined patterns of primary care use preceding a first diagnosis of psychotic disorder.

Methods:

We used health administrative data from Ontario (Canada) to construct a population-based retrospective cohort. We investigated patterns of primary care use, including frequency and timing of contacts, in the 6 years prior to a first diagnosis of psychosis, relative to a general population comparison group matched on age, sex, geographic area, and index date. We used latent class growth modeling to identify distinct trajectories of primary care service use, and associated factors, preceding the first diagnosis.

Results:

People with early psychosis contacted primary care over twice as frequently in the 6 years preceding first diagnosis (RR = 2.22; 95% CI, = 2.19 to 2.25), relative to the general population, with a sharp increase in contacts 10 months prior to diagnosis. They had higher contact frequency across nearly all diagnostic codes, including mental health, physical health, and preventative health. We identified 3 distinct service use trajectories: low-, medium-, and high-increasing usage.

Discussion:

We found elevated patterns of primary care service use prior to first diagnosis of psychotic disorder, suggesting that initiatives to support family physicians in their role on the pathway to care are warranted. Earlier intervention has implications for improved social, educational, and professional development in young people with first-episode psychosis.

Keywords

first-episode psychosis early intervention help-seeking pathways to care primary care family physicians health administrative data

Background

The prodrome to psychotic disorder is the period that begins with a deviation from a person’s previous behavior and extends to the onset of overt psychosis.^1,2 This prodromal phase can range from several weeks up to several years in duration.³ It is determined retrospectively after a first diagnosis of psychosis and is characterized by the presence of heterogeneous and nonspecific symptoms including depression, anxiety, social withdrawal, and attenuated positive and negative symptoms.⁴ This suggests that there may be help-seeking for mental health problems several years prior to the first diagnosis of psychotic disorder—indeed, prior research suggests that people with psychotic disorders were in contact with the mental health system for a median of 6 years preceding first diagnosis,⁵ with missed opportunities for early detection and intervention.³ Efforts to diagnose and minimize the duration of untreated psychosis during the early phase of illness—referred to herein as early psychosis—are essential for improving long-term outcomes.^6

–10

Pathways to care are the modes by which patients with mental health problems access help, including the help-seeking behavior of both the patient and family, the accessibility of mental health services, and the identification of and response to symptoms by each contact on the pathway to care.¹¹ For young people in the early stages of psychotic illness, pathways to care are complex, with multiple diverse contacts and associated treatment delays.^11,12 Young people with early psychosis often make multiple help-seeking attempts across providers and may experience negative and aversive pathways to care involving the criminal justice system, emergency department (ED), or inpatient unit.^13
–15 These contacts are associated with poor patient experience, disengagement from services, and high costs.¹⁴

We know that collectively, family physicians play an important role in the pathways to care for people with early psychosis.¹⁶ We have previously shown that nearly two thirds of young people with early psychosis make help-seeking attempts in primary care in the 6-month period preceding first diagnosis, and 1 in 3 people receive their first diagnosis from a family physician.¹⁷ Involvement of a family physician on the pathway to care reduces negative and aversive service contacts (i.e., police, ambulance, and the ED) and inpatient admission^14,16,18,19 and therefore is beneficial for improving service-related outcomes.²⁰ However, early psychosis is still a relatively infrequent occurrence in primary care, due to the low incidence of psychotic disorders,²¹ and family physicians estimate that they may only see 1 or 2 patients per year with the signs and symptoms of first-episode psychosis.²² Consequently, the importance of the family physician in early detection and intervention for psychotic disorder has gone largely unrecognized, and there is a need for early psychosis intervention services to more actively engage with the primary care sector.²³ However, there is a notable lack of literature on patterns of primary care use for young people with first-episode psychosis. Of exception, 2 prior studies have used health administrative data to look at help-seeking from a family physician before the first diagnosis of psychosis, showing that people with psychotic disorders have elevated contact rates in primary care up to 6 years prior to first diagnosis²⁴ and that consultation patterns may be used to identify high-risk subgroups who might benefit from further assessment and monitoring.²⁵ Because these patterns of service use will be highly dependent on the health system context, there is a need for Canadian data on this topic to highlight the important role of Canadian family physicians in pathways to care and identify opportunities for early psychosis intervention.

The objective of this study was to explore patterns of service use in primary care preceding a first diagnosis of nonaffective psychotic disorder. Using data from the provincial public health insurance plan in Ontario, Canada, we sought to compare the frequency and timing of primary care visits, and associated diagnostic categories, among people with early psychosis, relative to a matched comparison group from the general population. We also sought to identify distinct trajectories of service use in primary care and associated factors. This study uses a similar approach as Nørgaard and colleagues,²⁴ who analyzed primary care contacts among people with schizophrenia in Denmark, with additional analyses of the timing and trajectories of service use.

Methods

Study Design and Data Sources

We used population-based health administrative data to construct a retrospective cohort of people aged 14 to 35 years with newly diagnosed nonaffective psychotic disorder between April 2005 and March 2015 and a matched comparison group from the general population. We chose this age group because it is considered a priority population for early intervention by the Ontario Ministry of Health and Long-Term Care.²⁶ The linked databases are detailed in Online Appendix A and included information on basic sociodemographic characteristics, physician billings, hospitalizations, and ED visits. The data sets were linked using unique encoded identifiers and analyzed at ICES, which is an independent, nonprofit research institute whose legal status under Ontario’s health information privacy law allows it to collect and analyze health care and demographic data, without consent, for health system evaluation and improvement. We followed the Reporting of studies Conducted using Observational Routinely-collected Data (RECORD) guidelines for studies using health administrative data (Online Appendix B).²⁷

This study was based primarily on data from the Ontario Health Insurance Plan (OHIP), which contains physician billing claims paid for by the provincial public health insurance plan. The database includes information from approximately 94% of physicians in Ontario but excludes physicians paid by salary or other alternative practice models who do not submit shadow billings.

Cohort Definition

New cases of nonaffective psychotic disorder (i.e., schizophrenia, schizoaffective disorder, schizophreniform disorder, or psychosis not otherwise specified [NOS]) were identified by either a primary discharge diagnosis of nonaffective psychosis from an inpatient admission or at least 2 physician billing claims or ED visits with a diagnostic code for nonaffective psychosis in any 12-month period. A modified version of this algorithm has been validated using medical charts,²⁸ and this approach has been previously used to compute the incidence of nonaffective psychotic disorders in Ontario.^21,29 The index date was defined as the discharge date for cases identified from hospitalizations or the date of the first physician visit for cases identified from outpatient billings. People were excluded if they were missing data for age and sex variables (<1%).

All members of the general population who were eligible for OHIP and had no record of a psychotic disorder diagnosis were randomly assigned an index date from the distribution of index dates among the cases. We randomly selected 4 comparisons per case from the general population, matched on age at index date, sex, geographic area (based on first 3 digits of the postal code), and index date (± 6 months), using the greedy method without replacement.³⁰

Variable Definitions

Contact with primary care was defined as a visit with a family physician or pediatrician in an outpatient setting. We extracted information on all contacts with primary care for each person in the study cohort for the 6-year period preceding the index date. We generated counts for the total number of contacts, the number of contacts per year, and the number of contacts bimonthly over the observation period. We also computed the number of contacts per diagnostic code category and created a binary indicator variable to identify those who contacted primary care at least once with a given diagnostic code category over the observation period. Diagnostic code categories were further subdivided as mental health, physical health, and preventative health.

Sociodemographic variables included age at index date, sex, and rural place of residence. Age was categorized into groups by ICES (Table 1) to mitigate reidentification risk. Clinical factors included the type of psychotic disorder diagnosis (cases only, classified as schizophrenia spectrum or psychosis NOS), presence of an aggregated diagnosis group (ADG) chronic medical condition, and presence of an ADG chronic psychosocial condition. The Johns Hopkins Adjusted Clinical Group® System Version 10³¹ categorizes illnesses into 32 diagnostic clusters, known as ADGs, with similar clinical criteria and expected need for health care resources based on International Classification of Disease codes, which are obtained from both physician billing claims and hospital discharge abstracts.^32,33 Finally, variables reflecting service use history in the previous 2 years were created to indicate whether each person had access to a regular family physician and whether each person had a psychiatric ED visit or hospitalization in the 2 years prior to the index date. All covariates were measured at the index date.

Table 1.

Sociodemographic, Clinical, and Service Use History Characteristics for Cases with Early Psychosis and the General Population Comparison Group.

Variables		Cases N (%)	Comparison Group N (%)	Standardized Difference
Sociodemographic characteristics
Age	14 to 17 years	5,317 (13.5)	21,268 (13.5)	Matched
	18 to 20 years	8,085 (20.5)	32,340 (20.5)
	21 to 23 years	6,963 (17.7)	27,852 (17.7)
	24 to 26 years	5,533 (14.0)	22,132 (14.0)
	27 to 29 years	4,893 (12.4)	19,572 (12.4)
	30 to 32 years	4,380 (11.1)	17,520 (11.1)
	33 to 35 years	4,278 (10.8)	17,112 (10.8)
Sex	Male	25,049 (63.5)	100,196 (63.5)
Sex	Female	14,400 (36.5)	57,600 (36.5)
Location of residence	Urban	35,920 (91.1)	143,818 (91.1)
Location of residence	Rural	3,515 (8.9)	13,935 (8.8)
Clinical characteristics
Index diagnosis	Schizophrenia spectrum	19,408 (49)	N/A	N/A
Index diagnosis	Psychosis NOS	20,041 (51)	N/A	N/A
Number of ADGs	1 (Low, < 5)	18,688 (47.4)	126,788 (80.4)	0.73
	2 (Medium, 6 to 9)	13,190 (33.4)	25,025 (15.9)	0.42
	3 (High, 10 +)	7,571 (19.2)	5,983 (3.8)	0.50
Chronic medical condition	Yes	11,530 (29.2)	23,646 (15.0)	0.35
Chronic psychosocial condition	Yes	31,071 (78.8)	25,096 (15.9)	1.62
Service use history characteristics
Psychiatric hospitalization in 2 years prior to index date	Yes	4,749 (12.0)	589 (0.4)	0.50
Any hospitalization in 2 years prior to index date	Yes	3,947 (10.0)	8,019 (5.1)	0.19
Psychiatric ED visit in 2 years prior to index date	Yes	13,037 (33.1)	2,672 (1.7)	0.91
Any ED visit in 2 years prior to index date	Yes	22,669 (57.5)	43,266 (27.4)	0.64
Regular family physician	Yes	28,262 (71.6)	88,608 (56.3)	0.33

Note. ADG = aggregated diagnosis groups; NOS = not otherwise specified; ED = emergency department.

Statistical Analyses

Analyses were conducted using SAS (Version 9.4) and R (Version 3.3.0). We used frequencies and proportions, as well as medians and interquartile ranges (IQR), to describe the sample and the number of primary care contacts for each group. Standardized differences were computed to examine differences between case and comparison groups³³—standardized differences greater than 0.10 were considered significant.³⁴

To compare the number of primary care contacts in the 6 years prior to the index date, we used a negative binomial regression model with robust variance estimators to account for the matched design. Among people with early psychosis, we examined the timing of changes in the frequency of primary care contacts using change-point analysis. Change-point analysis models the process of a sequence of observations undergoing sudden change at unknown times and to determine the point at which a change occurred.³⁵ We used change-point analysis to identify time points where changes occurred in the average number of contacts with primary care per person bimonthly prior to index date. We used the R changepoint package for analysis, and the cpt.meanvar function with binary segmentation (BinSeg) method was used to detect at least 1 change in the mean and/or variance of the data.³⁶ The time points in the sequence of observations where a change was detected and percentage of mean difference in change were calculated.

We compared reasons for primary care contacts between cases and comparisons using diagnostic codes associated with primary care contacts. We used unadjusted log-linked binomial models with robust variance estimators, stratified by sex, to estimate the relative risk (RR) and associated 95% confidence intervals (CI) of contacting primary care within each diagnostic category in the 6 years prior to the index date for people with early psychosis, relative to the matched comparison group. For models comparing the frequency and reasons for primary care contacts between cases and controls, we used unadjusted models because cases and their comparisons were matched on age at index date, sex, and location of residence and also because other clinical and service use covariates will impact the frequency of primary care service use, and including these variables in analyses may adjust away any effect.

Finally, we used latent class growth modeling (LCGM) to identify distinct service utilization trajectories among people with early psychosis in the 6 years preceding diagnosis. LCGM is used to identify distinct subgroups of people following a similar pattern of change over time on a given variable.³⁷ First, a linear trajectory was modeled for one subgroup, then additional subgroups were added until the adjusted model fit worsened. With the number of subgroups determined, the optimal trajectory shape (i.e., linear, quadratic, cubic) for each group was tested until the adjusted model fit worsened and nonsignificant polynomial terms were removed, resulting in the final model.³⁸ A detailed description of the model selection process and the statistical parameters of the model is available in Online Appendix C. Variables entered in the LCGM included the number of primary care contacts per year for each of the 6 years of observation. The number of primary care contacts was categorized into deciles based on the sample distribution (0, 1, 2 to 4, 5 to 7, 8 to 10, 11 to 14, 15 to 18, 19 to 25, 26 to 37, 38+ contacts). We again chose not to include additional covariates in the LCGM, so that our ability to meaningfully observe primary care service use in this population was not obstructed. Once the service use trajectories were identified, we used log-linked binomial models to examine sociodemographic, clinical, and service use factors associated with the high-use trajectory. These variables were chosen based on established evidence of their relationship with primary care service utilization, and there was no indication of multicollinearity.

Results

We identified 39,449 incident cases of nonaffective psychosis over the 10-year period, matched to 157,796 comparisons (Table 1). Of the identified cases, 59.5% (n = 23,469) made help-seeking attempts for mental health reasons in primary care prior to first diagnosis, whereas the remainder were diagnosed in secondary or tertiary care with no prior mental health contacts in primary care. Significant differences between people with early psychosis and the comparison group were found for clinical and service use indicators (Table 1).

Over the 6-year period leading up to first diagnosis, 97.3% (n = 38,379) of people with early psychosis had any primary care contact (for both mental health and nonmental health related visits), compared to 76.4% (n = 120,600) of the comparison group. The median number of contacts over this period was 19 (IQR = 10,36) for people with early psychosis and 8 (IQR = 1,18) for the comparison group (standardized difference = 0.82). People with early psychosis contacted primary care over twice as frequently in the 6 years leading up to first diagnosis, relative to the comparison group (RR = 2.22; 95% CI, 2.19 to 2.25). The median number of primary care contacts per year begins to increase approximately 2 years before diagnosis for people with early psychosis, with no change for the comparison group over the 6-year observation period (Online Appendix D).

Five change points were detected in the mean number of primary care contacts per person with early psychosis, specifically at 56, 46, 30, 20, and 10 months prior to first diagnosis (Figure 1). The percentage mean difference between change points increases slightly at each time point from 105% at the first to second change point to 113% at the third to fourth time point although these differences are small and unlikely to be of clinical significance. However, there is a large increase in the percentage mean difference (140%) from the fourth to fifth change point, which occurs at 10 months before first diagnosis.

Figure 1.

Mean number of primary care contacts per case with early psychosis, counted bimonthly over the 6 years leading up to first diagnosis (black line) and change point locations detected (red lines). Significant change points detected at 56, 46, 30, 20, and 10 months prior to first diagnosis.

Diagnostic Codes Associated with Primary Care Use

Both males and females with early psychosis have a higher prevalence of primary care contacts across nearly all diagnostic code categories, including mental health, physical health, and preventive health, compared to their matched comparisons (Table 2).

Table 2.

Proportions and RR for Diagnostic Code Categories associated with Primary Care Contacts for Male and Female Cases with Early Psychosis and the General Population Comparison Group, Who Contacted Primary Care At Least Once in the 6 Years Leading up to First Diagnosis, for Each Diagnostic Category.

Diagnostic Code Category	Cases	Comparisons	RR	95% CI
Diagnostic Code Category	%	%	RR	95% CI
Males
Mental health related	75.6	21.6	3.49	3.44 to 3.54
Physical health related	95.4	75.5	1.26	1.26 to 1.27
Preventative health related	31.2	24.3	1.29	2.26 to 1.31
Females
Mental health related	84.1	31.9	2.64	2.60 to 2.67
Physical health related	97.2	77.6	1.25	1.25 to 1.26
Preventative health related	65.9	53.5	1.23	1.21 to 1.25

Note. RR = risk ratio; CI = confidence interval.

The complete list of the contact prevalence and RR for all diagnostic code categories (where prevalence >5%) is presented in Tables 3 and 4, stratified by sex. Notably, both males and females with early psychosis were over 6 times more likely to contact with a diagnostic code related to substance-use disorder (males: RR = 6.11; 95% CI, 5.84 to 6.40; females: RR = 6.25; 95% CI, 5.82 to 6.72), and over 4 to 6 times more likely to contact with a diagnostic code related to a behavioral disorder (males: RR = 4.23; 95% CI, 4.02 to 4.45; females: RR = 6.00; 95% CI, 5.50 to 6.55), relative to the comparison group.

Table 3.

Proportions of Male Cases and Comparisons Who Contacted Primary Care At Least Once in the 6 Years Leading up to First Diagnosis of Psychotic Disorder for Each Diagnostic Code Category.

Diagnostic Code Category associated with Contact	Cases	Comparisons	RR	95% CI
Diagnostic Code Category associated with Contact	%	%	RR	95% CI
Mental health related
Mental health and addictions services	75.55	21.64	3.49	3.44 to 3.54
Nonpsychotic disorders	70.24	18.43	3.81	3.75 to 3.87
Substance use disorder	17.04	2.79	6.11	5.84 to 6.40
Behavior disorder	11.63	2.75	4.23	4.02 to 4.45
Habit spasms (i.e., tics, stuttering)	9.38	3.42	2.74	2.61 to 2.89
Social problems	8.85	2.58	3.43	3.25 to 3.63
Hyperkinetic syndrome	6.20	1.72	3.60	3.37 to 3.85
Physical health related
Nonmental health and addictions services	95.44	75.46	1.26	1.26 to 1.27
Diseases of the respiratory system	66.09	51.02	1.30	1.28 to 1.31
Accidents, poisonings, and violence	63.46	44.70	1.42	1.40 to 1.44
Diseases of the skin and subcutaneous tissue	50.20	37.01	1.36	1.34 to 1.38
Diseases of the nervous system and sense organs	44.96	30.41	1.48	1.45 to 1.50
Diseases of the musculoskeletal system and connective tissue	42.26	30.27	1.40	1.37 to 1.42
Symptoms, signs and ill-defined conditions	41.97	23.85	1.76	1.73 to 1.79
Infections and parasitic diseases	41.44	29.92	1.39	1.36 to 1.41
Diseases of the digestive system	40.67	24.24	1.68	1.65 to 1.71
Diseases of the circulatory system	24.41	14.77	1.65	1.61 to 1.70
Diseases of the genito-urinary system	17.31	10.24	1.69	1.64 to 1.75
Endocrine, metabolic, and immunity disorders	12.16	7.92	1.54	1.48 to 1.60
Preventative health related
Annual health examination (adolescents)	22.11	16.28	1.36	1.32 to 1.40
Immunization	10.81	9.40	1.15	1.10 to 1.20

Note. RR = risk ratio; CI = confidence interval.

Table 4.

Proportions of Female Cases and Comparisons Who Contacted Primary Care At Least Once in the 6 Years Leading up to First Diagnosis of Psychotic Disorder for Each Diagnostic Code Category.

Diagnostic Code Category associated with Contact	Cases	Comparisons	RR	95% CI
Diagnostic Code Category associated with Contact	%	%	RR	95% CI
Mental health related
Mental health and addictions services	84.12	31.89	2.64	2.60 to 2.67
Nonpsychotic disorders	81.14	29.10	2.79	2.75 to 2.83
Habit spasms (i.e., tics, stuttering)	12.65	5.18	2.44	2.31 to 2.58
Substance use disorder	12.26	1.96	6.25	5.82 to 6.72
Social problems	11.67	3.70	3.15	2.96 to 3.35
Behavior disorder	8.36	1.39	6.00	5.50 to 6.55
Physical health related
Nonmental health and addictions services	97.21	77.63	1.25	1.25 to 1.26
Diseases of the genito-urinary system	65.90	47.53	1.39	1.37 to 1.41
Diseases of the skin and subcutaneous tissue	59.26	43.71	1.36	1.33 to 1.38
Diseases of the nervous system and sense organs	58.97	39.64	1.49	1.46 to 1.51
Accidents, poisonings and violence	60.91	38.81	1.57	1.54 to 1.60
Infections and parasitic diseases	53.97	37.56	1.45	1.41 to 1.46
Diseases of the digestive system	58.90	36.82	1.60	1.57 to 1.63
Symptoms, signs and ill-defined conditions	54.25	33.00	1.64	1.61 to 1.68
Diseases of the musculoskeletal system and connective tissue	47.03	32.84	1.43	1.40 to 1.46
Diseases of the circulatory system	31.45	18.65	1.69	1.64 to 1.74
Complications of pregnancy, childbirth, and puerperium	23.29	18.60	1.25	1.21 to 1.30
Endocrine, metabolic, and immunity disorders	20.99	12.90	1.63	1.57 to 1.69
Diseases of blood and blood-forming organs	13.33	8.27	1.61	1.53 to 1.69
Neoplasms	10.76	7.76	1.39	1.31 to 1.46
Preventative health related
Family planning	45.88	38.00	1.21	1.18 to 1.23
Annual health examination	34.50	28.63	1.21	1.17 to 1.24
Immunization	15.33	13.07	1.17	1.12 to 1.23

Note. RR = risk ratio; CI = confidence interval.

Service Use Profiles

Among people with early psychosis, the number of primary care contacts per year was best fit by a 3-group model with linear, quadratic, and cubic terms. Bayesian Information Criterion for model selection and estimates of trajectory parameters are provided in Online Appendix C, and Online Appendix E shows the trajectory of each group. The first trajectory (n = 23,380; 59% of cases) identified people with low and steady primary care usage (second decile or 1 contact per year) with an increase to almost the third decile (2–4 contacts per year) approximately 1 year before first diagnosis. The second trajectory (n = 13,468; 34% of cases) identified people with medium and steady primary care usage (third decile or 2–4 contacts per year) with an increase to the fourth decile (5–7 contacts per year) approximately 1 year before first diagnosis. The third trajectory (n = 2,601; 7% of cases) identified people with high and increasing primary care usage, beginning at the fifth decile (8–10 contacts per year) at approximately 6 years before first diagnosis and steadily increasing each year to the seventh decile (15–18 contacts per year) at approximately 1 year before first diagnosis. These patterns of low-, medium-, and high-usage were similar for the comparison group (Online Appendix E) although at a lower frequency and without the increase in the final year before the index date. Additionally, the trajectory for the high-use group was not increasing over time in the comparison group.

We compared sociodemographic, clinical, and service use history characteristics for the high-increasing trajectory, relative to the other 2 trajectories combined (Table 5). People with early psychosis were more likely to be in the high-increasing-usage trajectory versus the low- and medium-usage trajectories, if they were older (aged 24+; RR = 2.55; 95% CI, 2.33 to 2.80), had a regular family physician (RR = 1.12; 95% CI, 1.01 to 1.24), had a chronic medical condition (RR = 4.10; 95%CI, 3.76 to 4.48), had a chronic psychosocial condition (RR = 6.70; 95% CI, 5.18 to 8.67), or had contact with acute care services in the 2 years prior to diagnosis (RR = 1.70; 95%CI, 1.56 to 1.85). People with early psychosis were less likely to be in the high-increasing-usage trajectory if they were subsequently diagnosed with psychosis NOS (vs. schizophrenia spectrum disorder; RR = 0.86; 95% CI, 0.79 to 0.93) or were male (RR = 0.38; 95% CI, 0.34 to 0.41).

Table 5.

Differences in Sociodemographic, Clinical, and Service Use History Characteristics between Early Psychosis Latent Class Growth Model Trajectory Groups.

Variables	Unadjusted RR	95% CI	Adjusted RR	95% CI
Psychosis NOS index diagnosis (ref. = schizophrenia spectrum)	0.80	0.74 to 0.86	0.86	0.79 to 0.93
Age at index date (ref. = youngest age group, 14–23)	2.85	2.62 to 3.10	2.55	2.33 to 2.80
Male sex	0.31	0.29 to 0.33	0.38	0.34 to 0.41
Access to regular family physician	1.49	1.36 to 1.63	1.12	1.01 to 1.24
Rural place of residence (ref. = urban or nonrural)	0.93	0.81 to 1.06	0.94	0.81 to 1.10
Chronic medical condition	4.94	4.56 to 5.34	4.10	3.76 to 4.48
Chronic psychosocial condition	10.86	8.47 to 13.93	6.70	5.18 to 8.67
Psychiatric ED visit or hospitalization in prior 2 years	2.18	2.03 to 2.35	1.70	1.56 to 1.85

Note. RR = risk ratio; CI = confidence interval; NOS = not otherwise specified; ref. = reference group; ED = emergency department.

Discussion

Our findings suggest that people with early psychosis contact primary care at a higher frequency and at an increasing rate, compared to the general population, during the 6 years preceding first diagnosis of a psychotic disorder. Our findings show repeated contacts with primary care prior to first diagnosis, which is consistent with other studies on first-episode psychosis.^24,25,39 It is important to note that this pattern may also be seen prior to first onset for other nonpsychiatric medical conditions.^40

–44

The trends we observed may reflect physician difficulty recognizing and responding to the emerging signs and symptoms of psychosis—including prodromal symptoms—possibly due to nonspecific or changing clinical presentation over time.⁴⁵ People with early psychosis use a range of mental health services for a wide array of psychiatric conditions prior to first diagnosis of psychosis,^5,16,46 and this generalized presentation often makes for a complex diagnostic process.^22,47

–50 Additionally, more than 1 consultation may be needed to gather sufficient information to make a diagnosis.^48
–50 Our findings may also suggest that the family physician may be conservative in assigning a diagnosis of psychotic disorder due to the gravity and consequences for the patient.⁵¹ These patterns may also reflect an absence of partnership with early psychosis intervention programs and other mental health services, as family physicians may have difficulty connecting the person with specialized care.^22,47

–50 These findings highlight opportunities for early intervention in the primary care context.

We found that males and females with early psychosis have higher visit frequencies not only for mental health–related diagnostic codes but also for physical and preventive health–related reasons. Previous research suggests that people with early psychosis are often in contact with primary care for both mental health reasons^5,16,46 and somatic reasons⁵² during the period before first diagnosis. The heterogeneity in diagnostic codes, including both mental health and physical health–related codes, are expected given the myriad of symptoms seen in prodromal psychosis, and are consistent with other studies using health administrative data to investigate prodromal help-seeking.^5,16 People with schizophrenia often have other physical comorbidities,^53

–59 and schizophrenia is associated with a range of preexisting somatic conditions, including infectious illness, autoimmune diseases, and head injury,^{52,60

–65} which may account for the physical health–related contacts that we observed. The higher number of contacts for preventative health–related diagnostic codes may be a result of family physicians practicing opportunistic prevention, where each contact provides an opportunity for the prevention of illness and encouragement of healthy lifestyles, and may involve the provision of additional services beyond the original presenting issue.⁶⁶

This study reports 3 trajectories of service use over the 6-year observation period for people with early psychosis: low-, medium-, and high-increasing usage. There is a lack of literature on service use trajectories in psychotic disorders; however, it may be that the trajectories identified are not unique to psychotic disorders and instead may be expected prior to first diagnosis for any serious medical condition.

Limitations

This study is limited by the variables available in the ICES data holdings, and we are missing information on important confounding factors, such as the severity of symptoms, acuity of first presentation, and ethnicity.^67

–71 We are unable to identify primary care visits initiated by parents or other caregivers who play an important role in help-seeking for early psychosis,⁷² and the patterns of primary care use may differ based on who initiates contact. The databases also do not have information on the timing of onset of psychotic symptoms, and thus, we cannot decipher between help-seeking for prodromal psychosis or active psychosis, which may confound the patterns observed. The ICES databases were not developed for research purposes, and the diagnostic codes assigned may not accurately represent the reason for contacting primary care, due to under- or overcoding of diagnoses and a lack of diagnostic standardization across professionals.^73,74 Unlike hospital data, physician billings are limited to 1 diagnostic code per contact; therefore, the diagnostic codes used may not provide an accurate description of the reason for the contact with primary care. Furthermore, when multiple contacts occurred on the same day, only the diagnostic code for the first contact was retained.

The algorithm used in this study was validated for chronic schizophrenia,²⁸ whereas the cases identified had a first onset of psychotic disorder; therefore, the algorithm performance may not be as stated in the original validation article. This study also defined the cohort using nonaffective psychosis, as we are unable to identify people with affective psychoses using health administrative data, and therefore, results do not generalize to all cases of first-episode psychosis. Our findings also cannot be extrapolated to older age groups who may have different patterns of help-seeking in primary care,⁷⁵ and our age restriction may have disproportionately excluded women with psychotic disorders who tend to have an older age at onset.⁷⁶ The data represent patterns of primary care use for cases of psychotic disorder diagnosed between 2005 and 2015, and ongoing changes to the mental health system in Ontario may have altered these patterns; therefore, updates to these analyses are warranted. Finally, the administrative data used in this study did not capture all contacts with primary care. The databases are missing information from nurse practitioners and salaried physicians (<6% of family physicians) who often provide primary care services in clinics for marginalized populations. Not all people in our sample had access to a regular family physician (72% for cases, 56% for comparison group), so we kept our definition of primary care broad to include walk-in clinics and other settings where family physicians may provide care. Nonetheless, some people may have been forced to turn to other service providers due to lack of access, meaning that help-seeking in primary care for early psychosis could be even higher if all young people had access to a regular family physician.

Conclusions

The patterns of primary care use we observed in this study support the notion that family physicians play an active role on the pathways to care for young people with early psychosis. We need to consider how we can better support family physicians in their role, perhaps through additional training and resource allocation to improve necessary collaborations between primary care and early psychosis intervention programs. Further research on this topic could incorporate information from electronic medical records and prescription databases to further delineate the practice patterns of family physicians treating people with early psychosis. Future research could also explore whether these patterns of service utilization in primary care could be used in risk prediction algorithms to distinguish nonaffective psychotic disorders from other mental health conditions. Increasing the detection of early psychosis in primary care could help reduce the duration of untreated psychosis and improve the social, educational, and professional outcomes in young people with first-episode psychosis.

Supplemental Material

Supplemental Material, FEP_Help-Seeking_in_Primary_Care_-_Appendices_Submission_#2 - Patterns of Primary Care Use Prior to a First Diagnosis of Nonaffective Psychotic Disorder in Ontario, Canada: Modèles d’utilisation des soins de première ligne avant un premier diagnostic de trouble psychotique non affectif en Ontario, Canada

Supplemental Material, FEP_Help-Seeking_in_Primary_Care_-_Appendices_Submission_#2 for Patterns of Primary Care Use Prior to a First Diagnosis of Nonaffective Psychotic Disorder in Ontario, Canada: Modèles d’utilisation des soins de première ligne avant un premier diagnostic de trouble psychotique non affectif en Ontario, Canada by Nicole Schoer, Rebecca Rodrigues, Jennifer Reid, Bridget L. Ryan, Daniel J. Lizotte, Richard Booth, Arlene G. MacDougall, Paul Kurdyak and Kelly K. Anderson in The Canadian Journal of Psychiatry

Footnotes

Authors’ Note

The project coinvestigators include Kelly K. Anderson (The University of Western Ontario); Suzanne Archie (McMaster University); Richard G. Booth (The University of Western Ontario); Elisa Candido (ICES); Chiachen Chang (Northern Ontario School of Medicine); Saadia Hameed (The University of Western Ontario); Liisa Jaakkimainen (ICES); Paul Kurdyak (Centre for Addiction and Mental Health); Daniel Lizotte (The University of Western Ontario); Arlene G. MacDougall (The University of Western Ontario); Ross Norman (The University of Western Ontario); Lena Palaniyappan (The University of Western Ontario); Rebecca Rodrigues (The University of Western Ontario); Bridget L. Ryan (The University of Western Ontario); Amanda L. Terry (The University of Western Ontario); Aristotle Voineskos (Centre for Addiction and Mental Health). Parts of this material are based on data and information compiled and provided by the Ontario Ministry of Health and Long-Term Care (MOHLTC), Cancer Care Ontario (CCO), and Canadian Institute for Health Information (CIHI). The analyses, conclusions, opinions and statements expressed herein are solely those of the authors and do not reflect those of the funding or data sources; no endorsement is intended or should be inferred. The data set from this study will be held securely in coded form at ICES and the ICES analyst will have full access to study data. While data sharing agreements prohibit ICES from making the data set publicly available, access can be granted to those who meet pre-specified criteria for confidential access, available at . The full dataset creation plan is available from the authors upon request.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The study was completed at the ICES Western site, where core funding is provided by the Academic Medical Organization of Southwestern Ontario, the Schulich School of Medicine and Dentistry, Western University, and the Lawson Health Research Institute. This study also received funding from a project grant from the Canadian Institutes of Health Research (PJT 153022).

ORCID iDs

Paul Kurdyak

Kelly K. Anderson

Source of Data and Ethics Approval

In 2018, the institute formerly known as the Institute for Clinical Evaluative Sciences formally adopted the initialism ICES as its official name. ICES is a prescribed entity under section 45 of Ontario’s Personal Health Information Protection Act (PHIPA). Section 45 is the provision that enables analysis and compilation of statistical information related to the management, evaluation and monitoring of, allocation of resources to, and planning for the health system. Section 45 authorizes health information custodians to disclose personal health information to a prescribed entity, like ICES, without consent for such purposes. The portions of this protocol that involve health administrative and electronic medical record data do not require review by a Research Ethics Board.

Supplemental Material

Supplemental material for this article is available online.

References

Yung

Mcgorry

. The prodromal phase of first-episode psychosis: past and current conceptualizations. Schizophr Bull. 1996;22(2):353–370.

Cornblatt

Lencz

Obuchowski

. The schizophrenia prodrome: treatment and high-risk perspectives. Schizophr Res. 2002;54(1-2):177–186.

Schultze-Lutter

Rahman

Ruhrmann

, et al. Duration of unspecific prodromal and clinical high risk states, and early help-seeking in first-admission psychosis patients. Soc Psychiatry Psychiatr Epidemiol. 2015;50(12):1831–1841.

Iyer

S. N

Boekestyn

Cassidy

Joober

Malla

. Signs and symptoms in the pre-psychotic phase: description and implications for diagnostic trajectories. Psychol Med. 2008;38(8):1147–1156.

Rietdijk

Hogerzeil

van Hemert

Cuijpers

Linszen

van der Gaag

. Pathways to psychosis: help-seeking behavior in the prodromal phase. Schizophr Res. 2011;132(2-3):213–219.

Perkins

Boteva

Lieberman

. Relationship between duration of untreated psychosis and outcome in first-episode schizophrenia: a critical review and meta-analysis. Am J Psychiatry. 2005; 162(10):1785–1804.

Marshall

Lewis

Lockwood

Drake

Croudace

. Association between duration of untreated psychosis and outcome in cohort of first-episode patients: a systematic review. Arch Gen Psychiatry. 2005;62(9):975–983.

Norman

RMG

Lewis

Marshall

. Duration of untreated psychosis and its relationship to clinical outcome. Br J Psychiatry. 2005;187(S48):s19–s23.

Mcgorry

Killackey

Yung

. Early intervention in psychotic disorders: detection and treatment of the first episode and the critical early stages. Med J Aust. 2007;187(S7):S8–S10.

10.

Murphy

Brewer

. Early intervention in psychosis: strengths and limitations of services. Adv Psychiatr Treat. 2011;17(6):401–407.

11.

Singh

Grange

. Measuring pathways to care in first-episode psychosis: a systematic review. Schizophr Res. 2006;81(1):75–82.

12.

MacDonald

Fainman-Adelman

Anderson

Iyer

. Pathways to mental health services for young people: a systematic review. Soc Psychiatry Psychiatr Epidemiol. 2018;53(10):1005–1038.

13.

Anderson

Fuhrer

Malla

. “There are too many steps before you get to where you need to be”: help-seeking by patients with first-episode psychosis. J Ment Heal. 2013;22:384–395.

14.

Anderson

Fuhrer

Schmitz

Malla

. Determinants of negative pathways to care and their impact on service disengagement in first-episode psychosis. Soc Psychiatry Psychiatr Epidemiol. 2013;48(1):125–136.

15.

von Reventlow

Krüger-Özgürdal

Ruhrmann

, et al. Pathways to care in subjects at high risk for psychotic disorders—a European perspective. Schizophr Res. 2014;152(2-3):400–407.

16.

Anderson

Fuhrer

Wynant

Abrahamowicz

Buckeridge

. Patterns of health services use prior to a first diagnosis of psychosis: the importance of primary care. Soc Psychiatry Psychiatr Epidemiol. 2013;48(9):1389–1398.

17.

Anderson

Kurdyak

. Factors associated with timely physician follow-up after a first diagnosis of psychotic disorder. Can J Psychiatry. 2017;62(4):268–277.

18.

Anderson

Flora

Ferrari

, et al. Pathways to first-episode care for psychosis in African-, Caribbean-, and European-origin groups in Ontario. Can J psychiatry. 2015;60(5):223–231.

19.

Power

Iacoponi

Reynolds

, et al. The Lambeth early onset crisis assessment team study: general practitioner education and access to an early detection team in first-episode psychosis. Br J Psychiatry. 2007;191(51):s133–s139.

20.

Anderson

Archie

Booth

, et al. Understanding the role of the family physician in early psychosis intervention. BJPsych Open. 2018;4(6):447–453.

21.

Anderson

Norman

MacDougall

, et al. Estimating the incidence of first-episode psychosis using population-based health administrative data to inform early psychosis intervention services. Psychol. Med. 2019;49(12):2091–2099.

22.

Simon

Lester

Tait

, et al. The International Study on General Practitioners and Early Psychosis (IGPS). Schizophr Res. 2009;108(1-3):182–190.

23.

Shiers

Lester

. Early intervention for first-episode psychosis. BMJ .2004;328(7454):1451–1452.

24.

Nørgaard

HCB

Pedersen

Fenger-Grøn

, et al. Increased use of primary care during 6 years of prodromal schizophrenia. Acta Psychiatr Scand. 2016;134(3):225–233.

25.

Sullivan

Hamilton

Tilling

, et al. Association of primary care consultation patterns with early signs and symptoms of psychosis. JAMA Netw Open. 2018;1(7):e185174.

26.

Ontario Ministry of Health and Long Term-Care. Program policy framework for early intervention in psychosis. Toronto: Ontario Ministry of Health and Long-Term Care. 2004.

27.

Benchimol

Smeeth

Guttmann

, et al. The REporting of studies conducted using observational routinely-collected health data (RECORD) statement. PLos Med. 2015;12(10):e1001885.

28.

Kurdyak

Lin

Green

Vigod

. Validation of a population-based algorithm to detect chronic psychotic illness. Can J Psychiatry. 2015;60(8):362–368.

29.

Anderson

Cheng

Susser

McKenzie

Kurdyak

. Incidence of psychotic disorders among first-generation immigrants and refugees in Ontario. CMAJ. 2015;187(9):E279–E286.

30.

Austin

. A comparison of 12 algorithms for matching on the propensity score. Stat Med. 2014;33(6):1057–1069.

31.

The Johns Hopkins University Bloomberg School of Public Health, H. S. R. & D. C. The Johns Hopkins ACG® Case-Mix System Version 10.0 Release Notes. (2011).

32.

Austin

Newman

Kurdyak

. Using the Johns Hopkins aggregated diagnosis groups (ADGs) to predict mortality in a population-based cohort of adults with schizophrenia in Ontario, Canada. Psychiatry Res. 2012;196(1):32–37.

33.

Yang

Dalton

. A unified approach to measuring the effect size between two groups using SAS. SAS Glob Forum. 2012;1–6.

34.

Austin

. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38(6):1228–1234.

35.

Barry

Hartigan

. A Bayesian analysis for change point problems. J Am Stat Assoc. 1993;88(42):309–319.

36.

Killick

Eckley

changepoint: an R package for changepoint analysis. Lancaster Univ. 2013;58(3):1–15.

37.

Andruff

Carraro

Thompson

Gaudreau

Louvet

. Latent class growth modelling: a tutorial. Tutor Quant Methods Psychol. 2009;5(1):11–24.

38.

Nagin

. Group-based modeling of development. Cambridge (MA): Harvard University Press, 2005.

39.

Chen

Farooq

Edwards

, et al. Patterns of symptoms before a diagnosis of first episode psychosis: a latent class analysis of UK primary care electronic health records. BMC Med. 2019;17(1):227.

40.

Coté

Easler

, et al. Informative patterns of health-care utilization prior to the diagnosis of pancreatic ductal adenocarcinoma. Am J Epidemiol. 2017;186(8):944–951.

41.

Jensen

Petersen

Kirketerp-Møller

Poulsen

Andersen

. Progression of disease preceding lower extremity amputation in Denmark: a longitudinal registry study of diagnoses, use of medication and healthcare services 14 years prior to amputation. BMJ Open. 2017;7(11):e016030.

42.

Rædkjær

Maretty-Kongstad

Baad-Hansen

, et al. Use of healthcare services two years before diagnosis in Danish sarcoma patients, 2000–2013. Sarcoma. 2019;2019:8108590.

43.

Akazawa

Halpern

Riedel

, et al. Economic burden prior to COPD diagnosis: a matched case-control study in the United States. Respir Med. 2008;102(12):1744–1752.

44.

Nygaard

Jensen

Christensen

Vedsted

Health care use before a diagnosis of primary intracranial tumor: a Danish nationwide register study. Clin Epidemiol. 2018;10:809–829.

45.

McGorry

Killackey

Yung

. Early intervention in psychosis: concepts, evidence and future directions. World Psychiatry. 2008;7(3):148–156.

46.

Simon

Stewart

Hunkeler

, et al. Care pathways before first diagnosis of a psychotic disorder in adolescents and young adults. Am J Psychiatry. 2018;175(5):434–442.

47.

Platz

Umbricht

Cattapan-Ludewig

, et al. Help-seeking pathways in early psychosis. Soc Psychiatry Psychiatr Epidemiol. 2006;41(12):967–974.

48.

Lester

Tritter

Sorohan

. Patients’ and health professionals’ views on primary care for people with serious mental illness: focus group study. Br Med J. 2005;330(7500):1122. doi: 10.1136/bmj.38440.418426.8F.

49.

Nkire

Sardinha’s

Nwosu’s

, et al. Evaluation of knowledge and attitudes among primary care physicians in Cavan–Monaghan as “gatekeepers-in-waiting” for the introduction of Carepath for overcoming psychosis early (COPE). Early Interv Psychiatry. 2015;9(2):141–150.

50.

Giannopoulos

Carroll

Ebmeier

. Improving the detection and treatment of schizophrenia. Practitioner. 2014;258(1771):11–16.

51.

Thomas

Seebohm

Wallcraft

Kalathil

Fernando

. Personal consequences of the diagnosis of schizophrenia: a preliminary report from the inquiry into the schizophrenia label. Ment Heal Soc Incl. 2013;17(3):135–139.

52.

Sørensen

Nielsen

Benros

Pedersen

Mortensen

.Somatic diseases and conditions before the first diagnosis of schizophrenia: a nationwide population-based cohort study in more than 900 000 individuals. Schizophr Bull. 2015;41(2):513–521.

53.

Laursen

Munk-Olsen

Vestergaard

. Life expectancy and cardiovascular mortality in persons with schizophrenia. Curr Opin Psychiatry. 2012;25(2):83–88.

54.

Pompili

Serafini

Innamorati

, et al. Suicide risk in first episode psychosis: a selective review of the current literature. Schizophr Res. 2011;129(1):1–11.

55.

The Schizophrenia Commission. The abandoned illness: a report by the Shizophrenia Commission. London, UK: The Schizophrenia Commission. 2012.

56.

Allebeck

. Schizophrenia: a life-shortening disease. Schizophr Bull. 1989;15(1):81–89.

57.

Zaheer

Jacob

de Oliveira

Rudoler

Juda

Kurdyak

. Service utilization and suicide among people with schizophrenia spectrum disorders. Schizophr Res. 2018;202:347–353.

58.

Coentre

Talina

Góis

Figueira

. Depressive symptoms and suicidal behavior after first-episode psychosis: a comprehensive systematic review. Psychiatry Res. 2017;253:240–248.

59.

Hellemose

LAA

Laursen

Larsen

Toender

. Accidental deaths among persons with schizophrenia: a nationwide population-based cohort study. Schizophr Res. 2018;199:149–153.

60.

Arias

Sorlozano

Villegas

, et al. Infectious agents associated with schizophrenia: a meta-analysis. Schizophr Res. 2012;136(1-3):128–136.

61.

Khandaker

Zimbron

Dalman

Lewis

Jones

PB.

Childhood infection and adult schizophrenia: a meta-analysis of population-based studies. Schizophr Res. 2012:139(1-3):161–168.

62.

Benros

Nielsen

Nordentoft

Eaton

Dalton

Mortensen

. Autoimmune diseases and severe infections as risk factors for schizophrenia: a 30-year population-based register study. Am J Psychiatry. 2011;168(12):1303–1310.

63.

Eaton

Byrne

Ewald

, et al. Association of schizophrenia and autoimmune diseases: linkage of Danish national registers. Am J Psychiatry. 2006;163(3):521–528.

64.

Orlovska

Byrne

Ewald

, et al. Head injury as risk factor for psychiatric disorders: a nationwide register-based follow-up study of 113,906 persons with head injury. Am J Psychiatry. 2014;171(4):463–469.

65.

Nielsen

Benros

Mortensen

. Hospital contacts with infection and risk of schizophrenia: a population-based cohort study with linkage of Danish national registers. Schizophr Bull. 2014;40(6):1526–1532.

66.

Turabian

Opportunistic prevention in family medicine: anticipatory care, case-finding and continuity of care. J Heal Care Prev. 2017;1(1):041–047.

67.

Rotenberg

Tuck

Ptashny

McKenzie

. The role of ethnicity in pathways to emergency psychiatric services for clients with psychosis. BMC Psychiatry. 2017;17:1–11.

68.

Bhui

Stephen

Sally

Stefan

Funke

Gene

. Ethnic variations in pathways to and use of specialist mental health services in the UK: systematic review. Br J Psychiatry. 2003;182:105–116.

69.

Anderson

Flora

Archie

Morgan

McKenzie

Race, ethnicity, and the duration of untreated psychosis: a systematic review. Soc Psychiatry Psych Epidemiol. 2014;49(7):1161–1174.

70.

Anderson

Flora

Archie

Morgan

Mckenzie

. A meta-analysis of ethnic differences in pathways to care at the first episode of psychosis. Acta Psychiatr Scand. 2014;130(4):257–268.

71.

Ferrari

Nina

Kelly

, et al. The African, Caribbean and European (ace) pathways to care study: a qualitative exploration of similarities and differences between African-origin, Caribbean-origin and European-origin groups in pathways to care for psychosis. BMJ Open. 2015;5(1):1–12.

72.

Chen

Gearing

Devylder

. Pathway model of parental help seeking for adolescents experiencing first-episode psychosis. Early Inter Psych. 2016;10(2):122–128.

73.

Lucyk

Tang

Quan

Barriers to data quality resulting from the process of coding health information to administrative data: a qualitative study. BMC Health Serv Res. 2017;17(1):1–10.

74.

Tang

Lucyk

Quan

. Coder perspectives on physician-related barriers to producing high-quality administrative data: a qualitative study. C Open. 2017;5(3):E617–E622.

75.

Copeland

John

Chen-Pin

, et al. Patterns of primary care and mortality among patients with schizophrenia or diabetes: a cluster analysis approach to the retrospective study of healthcare utilization. BMC Health Serv Res. 2009;9:1–11.

76.

Lappin

Margaret

Peter

, et al. Outcomes following first-episode psychosis—why we should intervene early in all ages, not only in youth. Aust N Z J Psychiatry. 2016;50(11):1055–1063.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB