Design and setting

The Nor-Hand study is a hospital-based observational cohort study. The study is prospective, and follows 300 patients with hand OA over approximately 8 years with three planned examinations. Patients between the ages of 40–70 years are recruited into Nor-Hand via two channels. One channel is through the rheumatology outpatient clinic at Diakonhjemmet, where patients referred to the hospital are screened for eligibility. Patients are asked to participate in the study if the rheumatologist confirms the diagnosis of hand OA and, at the same time, excludes other differential diagnoses. Patients are similarly recruited to the study through the ‘OA school’ organised by the Department of Rheumatology at Diakonhjemmet Hospital. The ‘OA school’ is a 1-day multidisciplinary course, to which patients can be directly referred from their general practitioner. However, only a few patients are recruited through this channel, as most are recruited through the outpatient clinic. Both populations are identically screened according to the eligibility criteria. Two follow-up examinations of the participants are planned. We plan to repeat all investigations from the baseline examination at the follow-up. However, if we do not get the funding to repeat all examinations, we will, at a minimum, include conventional radiographs and patient-reported outcomes.

Patients

The study population consists of men and women between the ages of 40 and 70 years. However, due to the female predominance among patients with hand OA referred to specialist care, the participants in the study are mainly women. Their diagnosis of hand OA is proven either by ultrasound and/or clinical examination performed by a rheumatologist at the rheumatology outpatient clinic at Diakonhjemmet Hospital. All rheumatologists at the department are encouraged to recruit patients to the study. The rheumatologists can choose whether they examine for OA using clinical examination or ultrasound (or both). The time between screening examination can vary from a few days to several months depending on the patients’ availability and the fact that the screening started well ahead of the baseline examination (March 2015). In addition, all patients must be able to sign and understand an informed consent form. The inclusion criteria are summarised in box. At the time of screening, patients are excluded from the study if they meet any of the exclusion criteria (box).

Sample size

The number of patients included in this study is not determined by power calculations but is a pragmatically chosen number based on experience from previous studies.2 18 46 Based on available funding, it is found feasible to include 300 participants in the study.

Assessments

The patients are invited to a test evening, when most of the examinations, with the exception of CT, conventional radiography and MRI, are performed. One afternoon every week, a team of six trained medical students and one rheumatologist at Diakonhjemmet Hospital examine on average eight new patients. Every test evening lasts for 3–4 hours, and one medical student performs the medical assessment and the functional tests, another student performs the pain sensitisation tests, a third student performs the FOI, a fourth student performs the ultrasound of the upper extremities, a fifth student performs the ultrasound of the lower extremities, whereas a sixth student collects blood and urine samples for the biobank. Joint assessment is performed by the rheumatologist. The examiners are blinded for the other results. The participants are also asked to respond to questions about demographic and clinical factors, in addition to OA specific questionnaires. CT, conventional radiography and MRI are performed after every test evening. We aim to assess all 300 patients enrolled in the study with all the investigations described in this protocol (except for participants that have contraindications for some of the investigations, as specified under the description of each investigation).

Medical assessment

A trained medical student measures the height of the patient to the nearest millimetre in the standing position. The weight of the patient is measured in kilograms with one decimal precision, while the patient is barefoot and with minimal clothing. Hip circumference is measured around the greater trochanters, while waist circumference is measured midway between the lowest rib and the iliac crest, after the patient has taken a deep breath in and out. Measurements are taken to the nearest millimetre. The patient is then asked to rest for 5 min before the blood pressure is measured in the sitting position. The measurements are repeated until two consecutive systolic and diastolic pressures have a difference of less than 5 mm Hg. The last measurements of both systolic and diastolic blood pressures are noted. In addition, the heart rate is taken after 5 min of rest.

The patient is also asked to answer a self-administered comorbidity questionnaire.47 This questionnaire was developed by a panel of five physicians who chose 12 groups of the most frequent medical conditions in general practice. The language of the questionnaire is simplified so it can be easily understood, without any prior medical knowledge. Questions about treatment and impact of the conditions on daily activities are included in the questionnaire to quantify the severity of the diseases.47 Afterwards, the medical student goes over the questions together with the patient, checking that the comorbidity corresponds with the list of medications that the patients are asked to bring with them to the test evening.

A small hair sample is collected from the back of the patient’s head and will be sent to an international laboratory for quantification of mean cortisol levels. The mean cortisol level will be measured to explore associations between stress exposure and pain. The amount of hair collected is equivalent to less than the width of a pencil.

Joint assessment

One rheumatologist or one rheumatology resident examine the bilateral first carpometacarpal (CMC-1), first to fifth metacarpophalangeal (MCP), first interphalangeal (IP-1) and second to fifth proximal interphalangeal (PIP) and distal interphalangeal (DIP) joints for soft tissue swelling, bony enlargement and joint tenderness according to the EULAR handbook.48 Joint tenderness is assessed using the Doyle index.49 The overall hand OA disease activity based on these examinations is summarised on an NRS from 0 to 10. The rheumatologist/rheumatology resident perform the joint assessment only, and are blinded to all other data collected in the study.

In the foot, the tibiotalar, talonavicular, medial naviculocuneiform, intermedial naviculocuneiform, subtalar, first metatarsophalangeal (MTP) and first to fourth tarsometatarsal (TMT) joints are examined for the absence/presence of tenderness, soft tissue swelling and bony enlargement. The fifth TMT joint is not included in the examination because of challenges related to the performance of a reliable examination of the joint because of its anatomy.

Finally, the physician investigates whether the patient fulfils the American College of Rheumatology criteria for OA in hips, knees and hands.50–52 Due to the lack of radiographs of hips and knees, we will use ultrasound features instead of conventional radiographs to complete the questions on imaging findings in the knee and hip criteria. Conventional radiographs of hips and knees are not collected as an attempt to reduce the total patient burden. Moreover, studies from knee and hand OA have demonstrated superior sensitivity of ultrasound to detect osteophytes as compared with conventional radiographs.9 53 We are not aware of any studies comparing ultrasound and conventional radiographs in hip OA.

Functional tests

Grip strength is measured by Jamar dynamometer. The procedure is completed with the patient sitting in a chair with his/her elbows at a 90-degree angle without any support of his/her arms. First, the dominant hand is tested by squeezing the dynamometer as hard as possible. This is repeated twice with 15 s of rest between the attempts. The measurements are noted in kilograms with one decimal precision, and then repeated for the non-dominant hand.54

The fine motor skills are tested by the Moberg Pick-up test. In total, 12 small objects are spread on a table in front of the patient within a radius of 20–30 cm. First, the dominant hand is tested by instructing the patient to pick up the objects one-by-one as fast as possible, and place them in a small box. The procedure is then repeated with the non-dominant hand. The time from the touch of the first object to the placement of the last object in the box is measured, and the patient fails the test if he/she uses more than 300 s to pick up all the objects.55

Pain sensitisation tests

Temporal summation: A set of seven punctuate probes with fixed intensities and exerted forces of 8, 16, 32, 64, 128, 256 and 512 nM is used. During the investigation, the patient has his/her eyes closed, and the examiner starts with tapping the first probe against the left radioulnar joint. The patient rates the feeling of pain on the NRS from 0 to 10, where 0 is no pain and 10 is the worst pain imaginable. The examination is repeated with the following probes in their numerical order until the patient rates the pain at 4 or higher on the NRS (0–10). The probe that evoked the most pain is then tapped 10 more times with a pace of one tap per second at the same location, and the patient rates the pain of the first, fifth and tenth tap on the NRS.

Pressure pain threshold (PPT): The interphalangeal joint where the patient reports the most pain is tested with a hand-held digital algometer (FPIX25 Wagner), which is placed perpendicular to the dorsal side over the joint. The pressure of the algometer is increased with 0.5 kg/s decided by a metronome, while the patient is resting his/her hands on a table. When the pressure becomes painful, the patient should ask for the test to be stopped. The examination is repeated three times with the algometer placed at slightly different positions over the same joint with a pause of 30 s between the measurements. The PPT is the average of the measurements. The patient is asked if any of his/her finger joints is pain free. The whole procedure is then repeated on this asymptomatic joint as well as on several distant control points, including the trapezius muscle, the left distal radioulnar joint and the tibialis anterior muscle.

Conditioned pain modulation: A blood pressure cuff is placed around the right upper arm of the patient approximately 3 cm proximal to the antecubital fossa. The blood pressure cuff is inflated to 200 mm Hg, while the patient holds a mantle of 0.5 kg and flexes and extends the wrist 10 times. Meanwhile, the patient rates the pain in the forearm on a 0–10 NRS. If the patient reports pain below 4, he/she is asked to perform another five exercises and so on until the level of pain is 4 or above. The amount of exercises completed and the pain rating in the forearm are noted. A repetition of the PPT test is performed on the left radioulnar joint when blood pressure cuff is still inflated. The procedure is stopped if the blood pressure cuff has been inflated for more than 2 min without the patient reporting pain.

Light touch: Two von Frey filaments with a strength of 2 g and 26 g are each used to touch the patient four times at the left distal radioulnar joint, which is used as a control site, and afterwards used again on the finger joint where the patient reports the most pain. The patient has his/her eyes closed, and the filaments are pressed against the skin with enough pressure to bend them. The patient is asked to report when he/she feels the filament touching the skin, and grade on the 0–10 NRS if he/she finds it painful. Finally, the procedure is repeated with a safety pin. With every touch of the skin, the examiner ensures that the skin is only lightly touched and not penetrated by the safety pin. However, if the pin causes a slight bleeding, it is carefully sterilised before it is used on the next patient.

The pain sensitisation tests are selected based on what we believe is tolerable for the patients. The PPT tests may affect the microcirculation in that particular area for a few minutes afterwards, and the patients never undergo the FOI examination right after the pain tests. The patients always wait for at least 30 min before they undergo the FOI examination.

Ultrasound

Ultrasound of the bilateral hands, acromioclavicular joints, feet, knees and hips are performed. To ensure standardised investigations, one trained medical student performs the upper extremity examination and another medical student performs the lower extremity examination throughout the study. The medical students have been trained and supervised by experienced ultrasonographers (HBH, AM) who step in when the students are prohibited to attend. These ultrasonographers have 5–16 years of experience. Before the first test evening, the medical students were instructed in detail how to perform the ultrasound assessments. In addition, the student and the expert performed the assessments at the first three test evenings together to ensure agreement of the scoring. A reliability exercise is being performed, where the expert and the student separately and independently scores all the joint regions. The results from the reliability exercise will be presented in future papers.

A GE Logic S8 ultrasound machine is used for investigation of the upper extremity, while the lower extremity is investigated with a GE Logic E9 ultrasound machine. The same machines with fixed settings are used throughout the study to make the assessments as standardised as possible. The Doppler is optimised as previously recommended by Torp-Pedersen et al,56 and there are no upgrades on the machines during the study.

At the ultrasound investigation of the hands, the bilateral CMC-1, first to fifth MCP, IP-1 and the second to fifth PIP and DIP joints are longitudinally scanned from the radial to the ulnar dorsal side with the patient sitting opposite the investigator with his/her hands on a table. In addition, a transverse scanning is completed if the presence of pathology is uncertain. A scoring system for ultrasound features of hand OA made by a group of experts is used to investigate the joints.10 This scoring system includes synovial hypertrophy and/or effusion, power Doppler signals and osteophytes, all on semiquantitative scales (0–3). To improve the reliability of the scoring, atlases of previously collected representative images of osteophytes and synovitis are being used.9 57

Ultrasound scanning of the bilateral acromioclavicular joints are performed with the patient sitting with his/her shoulders in a neutral position. Both acromioclavicular joints are investigated with the probe placed over the joints in the long axis of the clavicle. Osteophytes are scored on a 0–3 scale, where 0 is normal, 1 is mild pathology, 2 is moderate pathology and 3 is severe pathology.

The hips are investigated with the patient lying on an examination bed in supine position with the hips and knees extended and the feet in neutral position. The anterior aspects of the proximal femur (head and neck) are evaluated for osteophytes and capsular height in a longitudinal view. Osteophytes are scored on a 0–3 scale of occurrence, where 0 is none, 1 is mild degree, 2 is medium degree and 3 is severe degree of osteophytes.58 Capsular height is recorded as the largest perpendicular distance (in millimetres) from the middle of the femoral neck to the capsule. If fluid is present, the distance will increase.

The bilateral knees are assessed when the patient is lying in supine position on an examination bed. Osteophytes in medial and lateral tibiofemoral joints are investigated and scored on a scale from 0 to 3 with the knees in an extended position.53 In addition, synovitis/effusion in the suprapatellar recess is scored on a 0–3 scale according to a previously developed ultrasound atlas.57 The probe is placed in the sagittal plane lateral of the midline while the patient still has his/her knees fully extended (with active extension to increase fluid collection in the suprapatellar recess). Finally, the cartilage height in the femoral sulcus is measured in millimetres with the knees maximally flexed. We only assess the thickness of the cartilage in the sulcus, from the top of the interphase to the top of the bone surface. The probe is then placed transversely and just proximal to the patella.

The ultrasound investigation of the feet includes the bilateral tibiotalar, talonavicular, first to third naviculocuneiform, first to fourth TMT, first MTP, IP-1, lateral subtalar, medial subtalar and the calcaneocuboid joints. The patient is lying in supine position with flexed knee and his/her foot resting on the examination bed. The fifth TMT joint is not included because of difficulties performing a reliable assessment of this joint because of its anatomy. The presence of osteophytes, grey scale synovitis (hypertrophy and/or effusion) and power Doppler signals are scored on a 0–3 scale. Finally, we investigate whether the spring ligament is intact.

Fluorescence optical imaging

The Xiralite scanner is used to examine the degree of disturbed microcirculation as a proxy for joint inflammation in both hands. It includes a scanner with LED lights, a computer and a camera taking pictures every second for 6 min. The patient sets his/her hands on a preformed hand rest, and is given intravenous fluorescence dye (ICG pulsion, 0.1 mg/kg of the body weight). Patients with poor liver function (transaminases above twice the upper reference limit), poor renal function (glomerular filtration rate below 40 mL/min), untreated hyperthyroidism (fT4 above 21 pmol/L and thyroid-stimulating hormone (TSH) below 0.5 mIE/L), or a known allergy to iodine or indocyanine are excluded from the FOI investigation. In addition, women who are pregnant or breastfeeding should be excluded, although it has not been relevant in the Nor-Hand study.

During examination, enhancement of the fluorescence dye occurs in the hands, and a trained PhD student will later score the intensity of enhancement in the different joints on a scale from 0 to 3. Grade 1 inflammation is red, grade 2 is intense red, while grade 3 is white on a red background. The examiner does the grading by looking at the composite image, which is derived from the first 240 images taken by the camera. In addition, an evaluation of three different phases will be performed. Phase 1 is right after the fluorescence dye is given, and studies have proposed that increased intensity in this phase represents high inflammation activity in the joint. Phase 2 is the period with fluorescence enhancement in the fingertips, and is shown to be the most sensitive to detect inflammation. In the third phase, signals from the fingertips are no longer visualised, but fluorescence signals from inflamed tissue may remain. Studies indicate that inflammation in the later phase is connected to increased capillary permeability.25 26 A trained medical student performs the investigation, and this student is blinded to all other results. One PhD student, who has not performed the FOI examination, will later score the images according to a standardised scoring system.25

Conventional radiographs

Bilateral frontal images of the hands are obtained with a posterior–anterior view (source to image-receptor distance (SID): 115 cm, exposure: 46 kVp and 2 mAs). The patient sits with both hands pronated and the palmar surfaces placed on the detector. A slight ulnar deviation of the hands ensures that the index fingers are extensions of the long axis of radius. A trained reader will evaluate the DIP, PIP, MCP and thumb base joints according to validated scoring systems, such as the Kellgren-Lawrence scale,59 60 the Osteoarthritis Research Society International atlas61 and the Verbruggen Veys anatomical phase scoring system.62

Furthermore, frontal images of the bilateral feet as well as oblique and side images of each foot are obtained (SID 115 cm, exposure: 52 kVp and 2 mAs). The patient is lying on an examination bed while the knees are bent so their feet can be placed straight down on the detector. The talonavicular joint, the medial, intermedial and lateral naviculocuneiform joints, the first to fourth TMT joints and the first MTP and IP joint will be evaluated for radiographic OA features.

MRI of dominant hand

Patients are scheduled for an MRI of their dominant hand at a private imaging centre (Volvat) in Oslo, Norway. The acquisition is performed with a Siemens Aera 1.5T MRI scanner (Germany) and a 16-channel hand/wrist coil with a field of view covering both the thumb base and the interphalangeal joints. Intravenous contrast (Dotarem 279.3 mg/mL, 0.2 mL/kg body weight) is given unless contraindications, for example, previous allergic reactions or reduced kidney function (glomerular filtration rate <40 mL/min). The patients are supinated with feet first and the dominant hand along their side. The details of the sequences are shown in table 3. The T1 volumetric interpolated breath-hold examination (VIBE) with water excitation sequences are reconstructed into axial, coronal and sagittal planes with 2 mm slice thickness. The axial planes of the carpus were defined according to carpal bones, whereas the axial planes of the fingers were perpendicular to the metacarpal bones in the coronal plane. The sagittal slices were perpendicular to the coronal slices. The MRIs will be scored according to validated scoring systems for the interphalangeal joints (Hand OA MRI scoring system; HOAMRIS)13 63 and thumb base joints (Thumb base OA MRI scoring system; TOMS).14 The MRIs will be read by future PhD students, supervised by a musculoskeletal radiologist (KF) and an experienced reader (IKH).

CT of dominant hand

CT is performed of the dominant hand using a 64-channel 750 HD Discovery Machine (General Electric, USA) (exposure: 120 kVp, 30 mAs). The patients are placed in ‘Superman’ position with the dominant arm straightened above their head and the non-dominant hand along their body. This position prevents unnecessary radiation of the head and internal organs. The radiation dose, that is, CT Dose Index Volume is 4,3 mGy. Scanning is performed with 0.625 mm thin slices, which are automatically reconstructed to 1 mm thick slices in axial, coronal and sagittal planes. Because there are no available hand OA scoring systems for CT scans, we must define the features and their grading before we start the scoring of images. This work will be based on previously developed scoring systems for MRI and conventional radiographs. The CT scans will be read by a future PhD student under supervision of a musculoskeletal radiologist (KF) and the project leader (IKH).

Blood tests/Biobank

In order to check the eligibility criteria, blood samples are collected to measure ESR, C reactive protein, ferritin, anti-CCP and rheumatoid factor at the screening examination. In addition, the glomerular filtration rate is measured at a maximum of 3 months before the contrast-enhanced MRI. Alanine amino transferase (ALAT), aspartate amino transferase (ASAT) and thyroid status with fT4 and TSH are measured before the FOI scanning. At the test evening, whole blood, plasma and serum, in addition to a urine sample, are collected and stored in a certified biobank, which consists of a freezer with a temperature of −70°C. These samples will later be used only for research.

Project timeline

April 2016−May 2017: Carry out the baseline examination.

2017–2020: Data analysis and submission of results from the baseline examination.

2020: Carry out the first follow-up examination.

2020–2023: Data analysis and submission of results from the first follow-up examination.

2024: Carry out the second follow-up examination.

2024–2027: Data analysis and submission of results from the second follow-up examination.

Statistical analysis

We will perform both parametric and non-parametric statistical analyses, relying on the distribution of variables. We plan to perform cross-sectional analyses when all participants have finished the baseline examination, and longitudinal analyses when participants have finished at least one of the two planned follow-up examinations. Reliability will be evaluated using, for example, kappa values and intraclass correlation coefficients. Agreement and associations between different relevant biomarkers and associations to patient-reported outcomes will be explored by, for example, regression analyses of our cross-sectional data. Analyses will be performed at patient level as well as joint level depending on our research question, and the dependency between joints within one patient will be taken into account. The selection of independent and dependent variables in our analyses will depend on the research question. Our analyses will be adjusted for multiple comparisons, if appropriate. Different hand OA subgroups will be compared. In longitudinal analyses, we will evaluate the predictive value of the baseline variables on hand OA outcomes as well as the sensitivity to change and the inter-relationship between OA biomarkers and patient-reported outcomes.