Morning Stiffness in Arthralgia and Early Arthritis
Morning Stiffness in Arthralgia and Early Arthritis
All datasets and cohorts used are described in more detail elsewhere.
In short, the arthralgia patients were referred to the Early Arthritis Recognition Clinics (EARC) from Leiden and Groningen (the Netherlands) or included in the Rotterdam Early Arthritis Cohort (REACH) (the Netherlands). The EARCs were initiated to reduce referral delay by GPs. It was observed that the GP delay contributed to two-thirds of the total delay between symptom onset and first visit to a rheumatologist in the Netherlands and that GPs frequently applied a 'wait-and-see' approach if they were unsure of the presence of arthritis. Therefore, GPs were instructed to refer to the EARC if they were undecided. The EARCs are early access clinics in which experienced rheumatologists screen patients on the presence of arthritis by physical examination; no laboratory investigations are done. The studied arthralgia patients visited the Leiden EARC between September 2010 and August 2013 and the Groningen EARC between October 2010 and January 2014. The EARC is primarily part of our care. Patients visiting the EARCs were not subjected to procedures that were done for scientific purposes, such as blood taking for biobanking. Therefore, in line with the Dutch law 'Wet medisch-wetenschappelijk onderzoek met mensen' (translated as 'the law on medical and scientific research involving people'), patients were not asked to sign an informed consent form.
The REACH study is an inception cohort that was initiated in the Rotterdam area, the Netherlands, in 2004. Inclusion required either pain or loss of movement in ≥2 joints or >1 swollen joint and ≥2 of the following items: unable to clench a fist in the morning, pain when shaking someone's hand, pins and needles in the fingers, difficulties wearing rings or shoes, a family history of RA, morning stiffness >1 hour, unexplained fatigue, all <1 year. The patients studied here were referred with joint symptoms by GPs between 2004 and 2009 and the presence of arthritis was assessed at the first visit. REACH was approved by the ethics committees of all three participating hospitals (Erasmus MC University Medical Center Rotterdam, Sint Franciscus Gasthuis Rotterdam and Maasstad Ziekenhuis Rotterdam); all patients gave written informed consent.
The early arthritis patients were included in the Leiden Early Arthritis Clinic (EAC) and the Evaluation et Suivi de POlyarthrites Indifférenciées Récentes (ESPOIR) cohort. The Leiden EAC is a population-based inception cohort that started in 1993. Inclusion required the presence of arthritis of ≥1 joint at physical examination and symptom duration <2 years. The patients studied were included between 1993 and 2011. The EAC was approved by the local Leiden University Medical Center (LUMC) ethics committee, all patients gave informed consent. ESPOIR is a nationwide cohort in which 14 rheumatology centres throughout France collaborate. Early arthritis patients are included if the treating rheumatologists suspected them of having or developing RA. Further, for inclusion patients had to be aged 18 to 70 years and to have ≥2 swollen joints for >6 weeks and <6 months. Patients studied were included between 2002 and 2005. ESPOIR was approved by the ethics committee of Montpellier; all patients gave written informed consent. In both the EAC and ESPOIR questionnaires were filled in, joint counts performed and laboratory evaluations done at baseline. Patients were followed prospectively with yearly follow-up visits; these included clinical and laboratory evaluations and radiographs of hands and feet.
In all cohorts the duration of morning stiffness was reported in minutes. In the EARCs, patients answered questionnaires on the presence and duration of morning stiffness (Additional file 1 http://www.arthritis-research.com/content/17/1/108/additional). In the EAC, ESPOIR and REACH the questions on presence and duration of morning stiffness were asked by trained research nurses (Additional file 1 http://www.arthritis-research.com/content/17/1/108/additional). Patients were not asked for specific locations of stiffness. Morning stiffness duration was dichotomized into <60 and ≥60 minutes. Sensitivity analyses were performed with ≥30 and ≥90 minutes as cutoffs. To evaluate the consistency in results when morning stiffness was assessed differently, analyses on arthritis patients were repeated with the severity of morning stiffness instead of the duration. The severity was assessed using a visual analogue scale (VAS) in 1,959 EAC patients included between 1993 and February 2010 and in all ESPOIR patients. For analyses, the VAS was divided into the three categories; mild 0 to 33 millimeter (mm), moderate 34 to 67 mm and severe 68 to 100 mm. In the arthralgia datasets the severity was not recorded. In all patients studied, morning stiffness was assessed at the first visit, when patients were not treated with disease-modifying antirheumatic drugs (DMARDs).
The outcomes were different in the three parts of this study (Figure 1).
(Enlarge Image)
Figure 1.
Outline of study questions. Of the 2010-RA patients in the EAC, radiographs were scored for the patients included between 1993 and 2006 (n = 636). Baseline characteristics of RA patients included before or after 2006 were not different. In ESPOIR, radiographic data was available for 659 of 677 RA patients. Here also, baseline characteristics of patients with and without radiographs were not different. EAC, Early Arthritis Clinic; ESPOIR, Evaluation et Suivi de POlyarthrites Indifférenciées Récentes; RA, rheumatoid arthritis.
Diagnostic Value in Arthralgia. In arthralgia patients, the outcome was the presence of arthritis ascertained at physical examination by experienced rheumatologists (assessed at the same visit when morning stiffness was evaluated). In both EARCs, a small proportion of patients (58 and 25) had no evident arthritis but were also not classified as having 'no arthritis' because the rheumatologists suspected these patients of RA development; these patients were excluded from analyses.
Diagnostic Value in Early Arthritis. In early arthritis, we aimed to assess the diagnostic value of morning stiffness and here the outcome was the presence of RA after one year. RA was defined as fulfilling the 2010 ACR/EULAR criteria for RA during the first year. An advantage of the 2010-RA criteria is that morning stiffness is not included, preventing circle reasoning. Since during the first weeks the diagnoses may not yet be definitive, the classification after year one was evaluated. These first two parts evaluated the diagnostic value.
Prognostic Value Within RA. Third, within 2010-RA patients, the prognostic value was assessed by studying two long-term outcomes. Structural damage was assessed using serial hands and feet radiographs that were scored according to the Sharp/van der Heijde (SHS) method with known time order and blinded to clinical data. In the EAC, radiographs were scored of patients included between 1993 and 2006. The follow-up was seven years in the EAC and three years in ESPOIR. The within-reader intraclass correlation coefficients (ICCs) were 0.91 and 0.87 for two readers in the EAC and 0.97 in ESPOIR. DMARD-free sustained remission, the opposite of disease persistence, was defined as the sustained absence of arthritis after discontinuation of DMARD therapy, including biologics and glucocorticoids (systemic and intra-articular), for the entire period of follow-up, which was at least one year. In the EAC, it was assessed by exploring the medical files until 10 years of follow-up. In ESPOIR, it was assessed over five years of follow-up by reviewing the structured visits in the ESPOIR database.
Characteristics were compared using Student t tests, Mann–Whitney tests or chi-square tests when appropriate. Associations of morning stiffness in arthralgia and early arthritis were done using logistic regression analyses. All analyses were adjusted for age and gender (although morning stiffness was not correlated with age in arthralgia or early arthritis). In early arthritis further adjustments were made for anti-citrullinated protein antibody (ACPA), rheumatoid factor (RF), swollen joint count (SJC), erythrocyte sedimentation rate (ESR) and symptom duration at baseline. The test characteristics (sensitivity, specificity), positive/negative predictive values (PPV/NPV) and area under the receiver operating characteristic curve (AUC) were calculated. This curve was used to derive the morning stiffness duration with the optimal discriminative ability (Youden's index). Associations between morning stiffness at baseline and radiographic progression were studied using multivariate normal regression analysis with log-transformed radiographic data as response variable as described elsewhere. Analyses on DMARD-free sustained remission were done by Kaplan-Meier survival curves and Cox proportional hazard regression models with morning stiffness as an independent variable. Analyses on radiographic progression and DMARD-free sustained remission were adjusted for age, gender, ACPA and inclusion period as a proxy for differences in treatment strategy as described elsewhere. SPSS version 20.0 was used (IBM Corp, Armonk, NY, USA).
Methods
Patients
All datasets and cohorts used are described in more detail elsewhere.
In short, the arthralgia patients were referred to the Early Arthritis Recognition Clinics (EARC) from Leiden and Groningen (the Netherlands) or included in the Rotterdam Early Arthritis Cohort (REACH) (the Netherlands). The EARCs were initiated to reduce referral delay by GPs. It was observed that the GP delay contributed to two-thirds of the total delay between symptom onset and first visit to a rheumatologist in the Netherlands and that GPs frequently applied a 'wait-and-see' approach if they were unsure of the presence of arthritis. Therefore, GPs were instructed to refer to the EARC if they were undecided. The EARCs are early access clinics in which experienced rheumatologists screen patients on the presence of arthritis by physical examination; no laboratory investigations are done. The studied arthralgia patients visited the Leiden EARC between September 2010 and August 2013 and the Groningen EARC between October 2010 and January 2014. The EARC is primarily part of our care. Patients visiting the EARCs were not subjected to procedures that were done for scientific purposes, such as blood taking for biobanking. Therefore, in line with the Dutch law 'Wet medisch-wetenschappelijk onderzoek met mensen' (translated as 'the law on medical and scientific research involving people'), patients were not asked to sign an informed consent form.
The REACH study is an inception cohort that was initiated in the Rotterdam area, the Netherlands, in 2004. Inclusion required either pain or loss of movement in ≥2 joints or >1 swollen joint and ≥2 of the following items: unable to clench a fist in the morning, pain when shaking someone's hand, pins and needles in the fingers, difficulties wearing rings or shoes, a family history of RA, morning stiffness >1 hour, unexplained fatigue, all <1 year. The patients studied here were referred with joint symptoms by GPs between 2004 and 2009 and the presence of arthritis was assessed at the first visit. REACH was approved by the ethics committees of all three participating hospitals (Erasmus MC University Medical Center Rotterdam, Sint Franciscus Gasthuis Rotterdam and Maasstad Ziekenhuis Rotterdam); all patients gave written informed consent.
The early arthritis patients were included in the Leiden Early Arthritis Clinic (EAC) and the Evaluation et Suivi de POlyarthrites Indifférenciées Récentes (ESPOIR) cohort. The Leiden EAC is a population-based inception cohort that started in 1993. Inclusion required the presence of arthritis of ≥1 joint at physical examination and symptom duration <2 years. The patients studied were included between 1993 and 2011. The EAC was approved by the local Leiden University Medical Center (LUMC) ethics committee, all patients gave informed consent. ESPOIR is a nationwide cohort in which 14 rheumatology centres throughout France collaborate. Early arthritis patients are included if the treating rheumatologists suspected them of having or developing RA. Further, for inclusion patients had to be aged 18 to 70 years and to have ≥2 swollen joints for >6 weeks and <6 months. Patients studied were included between 2002 and 2005. ESPOIR was approved by the ethics committee of Montpellier; all patients gave written informed consent. In both the EAC and ESPOIR questionnaires were filled in, joint counts performed and laboratory evaluations done at baseline. Patients were followed prospectively with yearly follow-up visits; these included clinical and laboratory evaluations and radiographs of hands and feet.
Assessment of Morning Stiffness
In all cohorts the duration of morning stiffness was reported in minutes. In the EARCs, patients answered questionnaires on the presence and duration of morning stiffness (Additional file 1 http://www.arthritis-research.com/content/17/1/108/additional). In the EAC, ESPOIR and REACH the questions on presence and duration of morning stiffness were asked by trained research nurses (Additional file 1 http://www.arthritis-research.com/content/17/1/108/additional). Patients were not asked for specific locations of stiffness. Morning stiffness duration was dichotomized into <60 and ≥60 minutes. Sensitivity analyses were performed with ≥30 and ≥90 minutes as cutoffs. To evaluate the consistency in results when morning stiffness was assessed differently, analyses on arthritis patients were repeated with the severity of morning stiffness instead of the duration. The severity was assessed using a visual analogue scale (VAS) in 1,959 EAC patients included between 1993 and February 2010 and in all ESPOIR patients. For analyses, the VAS was divided into the three categories; mild 0 to 33 millimeter (mm), moderate 34 to 67 mm and severe 68 to 100 mm. In the arthralgia datasets the severity was not recorded. In all patients studied, morning stiffness was assessed at the first visit, when patients were not treated with disease-modifying antirheumatic drugs (DMARDs).
Outcomes
The outcomes were different in the three parts of this study (Figure 1).
(Enlarge Image)
Figure 1.
Outline of study questions. Of the 2010-RA patients in the EAC, radiographs were scored for the patients included between 1993 and 2006 (n = 636). Baseline characteristics of RA patients included before or after 2006 were not different. In ESPOIR, radiographic data was available for 659 of 677 RA patients. Here also, baseline characteristics of patients with and without radiographs were not different. EAC, Early Arthritis Clinic; ESPOIR, Evaluation et Suivi de POlyarthrites Indifférenciées Récentes; RA, rheumatoid arthritis.
Diagnostic Value in Arthralgia. In arthralgia patients, the outcome was the presence of arthritis ascertained at physical examination by experienced rheumatologists (assessed at the same visit when morning stiffness was evaluated). In both EARCs, a small proportion of patients (58 and 25) had no evident arthritis but were also not classified as having 'no arthritis' because the rheumatologists suspected these patients of RA development; these patients were excluded from analyses.
Diagnostic Value in Early Arthritis. In early arthritis, we aimed to assess the diagnostic value of morning stiffness and here the outcome was the presence of RA after one year. RA was defined as fulfilling the 2010 ACR/EULAR criteria for RA during the first year. An advantage of the 2010-RA criteria is that morning stiffness is not included, preventing circle reasoning. Since during the first weeks the diagnoses may not yet be definitive, the classification after year one was evaluated. These first two parts evaluated the diagnostic value.
Prognostic Value Within RA. Third, within 2010-RA patients, the prognostic value was assessed by studying two long-term outcomes. Structural damage was assessed using serial hands and feet radiographs that were scored according to the Sharp/van der Heijde (SHS) method with known time order and blinded to clinical data. In the EAC, radiographs were scored of patients included between 1993 and 2006. The follow-up was seven years in the EAC and three years in ESPOIR. The within-reader intraclass correlation coefficients (ICCs) were 0.91 and 0.87 for two readers in the EAC and 0.97 in ESPOIR. DMARD-free sustained remission, the opposite of disease persistence, was defined as the sustained absence of arthritis after discontinuation of DMARD therapy, including biologics and glucocorticoids (systemic and intra-articular), for the entire period of follow-up, which was at least one year. In the EAC, it was assessed by exploring the medical files until 10 years of follow-up. In ESPOIR, it was assessed over five years of follow-up by reviewing the structured visits in the ESPOIR database.
Analyses
Characteristics were compared using Student t tests, Mann–Whitney tests or chi-square tests when appropriate. Associations of morning stiffness in arthralgia and early arthritis were done using logistic regression analyses. All analyses were adjusted for age and gender (although morning stiffness was not correlated with age in arthralgia or early arthritis). In early arthritis further adjustments were made for anti-citrullinated protein antibody (ACPA), rheumatoid factor (RF), swollen joint count (SJC), erythrocyte sedimentation rate (ESR) and symptom duration at baseline. The test characteristics (sensitivity, specificity), positive/negative predictive values (PPV/NPV) and area under the receiver operating characteristic curve (AUC) were calculated. This curve was used to derive the morning stiffness duration with the optimal discriminative ability (Youden's index). Associations between morning stiffness at baseline and radiographic progression were studied using multivariate normal regression analysis with log-transformed radiographic data as response variable as described elsewhere. Analyses on DMARD-free sustained remission were done by Kaplan-Meier survival curves and Cox proportional hazard regression models with morning stiffness as an independent variable. Analyses on radiographic progression and DMARD-free sustained remission were adjusted for age, gender, ACPA and inclusion period as a proxy for differences in treatment strategy as described elsewhere. SPSS version 20.0 was used (IBM Corp, Armonk, NY, USA).
Source...